Lessons from Qui Tam Litigation

Home , State Secrets Privilege, Tout (company)

COLUMBIA LAW REVIEW

VOL. 114 DECEMBER 2014 NO. 8

ARTICLE

PRIVATE ENFORCEMENT’S PATHWAYS: LESSONS FROM QUI TAM LITIGATION

David Freeman Engstrom*

How does making law through private lawsuits differ from making law by other means? That question is especially important where legislators deputize “private attorneys general” as statutory enforcers, from antitrust and securities to civil rights and consumer protection. Yet legal scholars have not offered sustained theoretical or empirical analysis of how private enforcement efforts evolve over the life of a regulatory regime and, in turn, shape the elaboration of legal mandates. This Article begins filling that gap by sketching a pair of competing accounts of “private enforcement’s pathways.” The Article then tests these accounts using original data drawn from a growing and increasingly controversial litigation regime that is fueling intense debate about the quantity, quality, and public control of private enforcement efforts: qui tam litigation under the False Claims Act (FCA). Examining some 6,000 qui tam lawsuits since 1986, the analysis rejects the claim, common among private enforcement’s critics, that the regime’s recent growth constitutes an inefficient “gold rush” of private enforcement activity. Yet the data also support the concern that entrepreneurial private enforcers will relentlessly press law’s boundaries, exploiting regulatory ambiguities in industry-wide lawsuits that public- minded prosecutors would reject, thus driving law down interpretive pathways it would not travel if enforcement were in purely public hands. Mapping private enforcement’s pathways in this way offers fresh perspective on longstanding debate about the merits and demerits of private enforcement of public law and the complex relationship between litigation and democracy, in the FCA context and beyond.

* Professor of Law, Stanford Law School. This Article and the broader project of which it is part have benefited from presentations at Stanford, Harvard, Columbia, Georgetown, Minnesota, NYU, Virginia, Northwestern, Vanderbilt, Illinois, and Penn law schools. Thanks to Jim Breen, Lisa Bressman, Steve Burbank, Tino Cuéllar, Nora Freeman Engstrom, Sean Farhang, Jonah Gelbach, Eric Havian, Peter Hutt, Scott Hemphill, Joan Krause, Dave Pozen, Amanda Rose, Robert Salcido, Shelley Slade, Kevin Stack, David Studdert, Claire Sylvia, Rob Vogel, and Melissa Wasserman for helpful feedback and to Scott Ganz for excellent research assistance.

1913 1914 COLUMBIA LAW REVIEW [Vol. 114:1913

INTRODUCTION ...... 1914 I. PRIVATE ENFORCEMENT’S PATHWAYS:THEORETICAL FRAMEWORK...... 1924 A. Is Private Enforcement Rational? Litigation Quantity and the Calibration Challenge...... 1924 B. Private Legal Innovation, Statutory “Drift,” and the Puzzle of Public Control...... 1931 II. THE CASE OF QUI TAM LITIGATION...... 1943 A. Qui Tam Basics...... 1943 B. The Case for Qui Tam as a Case Study ...... 1945 C. Data Overview ...... 1950 III. CHARTING QUI TAM’S RISE:ANATOMY OF A LITIGATION “EXPLOSION”...... 1951 A. Filings ...... 1951 B. Recoveries...... 1957 C. Per-Filing “Efficiency” ...... 1959 IV. QUI TAM’S CHANGING NATURE,PRIVATE LEGAL INNOVATION, AND THE PUZZLE OF PUBLIC CONTROL ...... 1964 A. Litigation Scale and Scope ...... 1964 B. Case Studies in Private Legal Innovation and Public Control ..1968 1. Private Legal Innovation and the Legislative Connection: Pharmaceutical Average Wholesale Price Claims...... 1968 2. Private Legal Innovation and the Bureaucratic Connection: Oil- and Gas-Royalty Claims ...... 1976 3. Private Legal Innovation and Regulatory Colonization: Healthcare-Kickback and Self-Referral Claims ...... 1981 C. DOJ Oversight: Delegation or Decoupling? ...... 1991 V. PRIVATE ENFORCEMENT’S PATHWAYS REVISITED...... 1996 A. FCA Implications...... 1996 B. Regulatory Choice Beyond the FCA...... 1998 C. Litigation and Democracy ...... 2002 CONCLUSION ...... 2005

INTRODUCTION On November 4, 2013, the United States Department of Justice (DOJ) announced yet another eye-popping settlement in a case brought under the qui tam provisions of the False Claims Act (FCA).1 Those

1. See Press Release, U.S. Dep’t of Justice, Johnson & Johnson to Pay More than $2.2 Billion to Resolve Criminal and Civil Investigations (Nov. 4, 2013), http://www.justice. gov/opa/pr/2013/November/13-ag-1170.html [hereinafter U.S. DOJ, Johnson & Johnson Release] (on file with the Columbia Law Review); see also Katie Thomas, J.&J. to Pay $2.2 Billion in Risperdal Settlement, N.Y. Times (Nov. 4, 2013), http://www.nytimes.com/2013 /11/05/business/johnson-johnson-to-settle-risperdal-improper-marketing-case.html (on 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1915 provisions authorize private individuals, dubbed “relators,” to sue on behalf of the United States for fraud in connection with federal programs and expenditures and to earn a bounty equal to a portion of any money returned to the federal fisc.2 This time, healthcare giant Johnson & Johnson (J&J) agreed to pay a whopping $2.2 billion to resolve a range of allegations, among them that the company marketed drugs for uses the Food and Drug Administration had not approved, thus facilitating illegal government reimbursement to healthcare providers under Medicare and Medicaid.3 The J&J whistleblowers, all former employees, have since divided $168 million between them for their work on the case.4 Cases like these—and there are now a number of them5—have helped make qui tam one of the fastest growing and most controversial litigation regimes today. The regime has expanded rapidly since Congress revived the FCA’s qui tam provisions in 1986, generating some 3,000 lawsuits and $20 billion in recoveries in the last five years alone— numbers that rival securities or antitrust suits over the same period.6 file with the Columbia Law Review) (describing Johnson & Johnson’s settlement with DOJ as third-largest pharmaceutical settlement in U.S. history). 2. False Claims Act (FCA), 31 U.S.C. §§ 3729–3733 (2012). 3. See U.S. DOJ, Johnson & Johnson Release, supra note 1 (announcing J&J settlement). Note that the settlement covered claims under both the federal FCA and state equivalents, with the federal portion accounting for roughly $800 million of the total. Id. 4. See Gregory Wallace, $168 Million Payout to Johnson & Johnson Whistleblowers, CNN Money (Nov. 4, 2013, 5:04 PM), http://money.cnn.com/2013/11/04/news/johnson- and-johnson-whistleblower-payout/index.html (on file with the Columbia Law Review) (discussing distribution of $168 million among J&J whistleblowers). 5. See Top 100 FCA Cases, Taxpayers Against Fraud Educ. Fund, http://www.taf. org/general-resources/top-100-fca-cases (on file with the Columbia Law Review) (last visited Sept. 15, 2014) (cataloguing “top 100” settlements under federal and state FCAs, including forty-four in excess of $200 million); see also infra Figure 4 (offering graphical summary of federal-level FCA recoveries since 1986). Another example of a J&J-like recovery is a July 2012 settlement with pharmaceutical giant GlaxoSmithKline for $3 billion, $2 billion of which resolved FCA allegations deriving from qui tam lawsuits. Press Release, U.S. Dep’t of Justice, GlaxoSmithKline to Plead Guilty and Pay $3 Billion to Resolve Fraud Allegations and Failure to Report Safety Data (July 2, 2012), http://www. justice.gov/opa/pr/2012/July/12-civ-842.html (on file with the Columbia Law Review); see also Katie Thomas & Michael S. Schmidt, Glaxo Agrees to Pay $3 Billion in Fraud Settlement, N.Y. Times (July 2, 2012), http://www.nytimes.com/2012/07/03/business/ glaxosmithkline-agrees-to-pay-3-billion-in-fraud-settlement.html (on file with the Columbia Law Review) (describing GlaxoSmithKline’s settlement with DOJ as largest settlement involving pharmaceutical company). 6. For aggregated statistics on qui tam filings and recoveries, see U.S. Dep’t of Justice, Fraud Statistics—Overview: Oct. 1, 1987–Sept. 30, 2013 (2013) [hereinafter Fraud Statistics], available at http://www.justice.gov/civil/docs_forms/C-FRAUDS_FCA_ Statistics.pdf (on file with the Columbia Law Review) (providing data on non–qui tam and qui tam new matters, settlements and judgments, and relator-share awards); see also Cornerstone Research, Securities Class Action Filings: 2012 Year in Review 3 (2013), available at http://www.cornerstone.com/getattachment/81e1daa9-7a2e-4589-9dd6-a4439 50b0837/Securities-Class-Action-Filings%E2%80%942012-Year-in-Revie.aspx (on file with 1916 COLUMBIA LAW REVIEW [Vol. 114:1913

Rising filings and big paydays have also spurred heated debate. Qui tam’s champions point to the regime’s rapid growth and its “spectacular results” as evidence of widespread corporate fraud on the government and the need for robust private enforcement to police it.7 Critics, by contrast, paint qui tam as a litigation regime run amok. The system, they complain, has produced a litigation “explosion” and a “gold rush” of lawsuits by disgruntled former employees and other assorted opportun- ists, including a growing qui tam plaintiffs’ bar, all seeking “jackpot” payouts.8 Worse, these critics continue, bounty-hunting privateers have the Columbia Law Review) (reporting 152 filings in 2012 and 188 filings in 2011); Ellen M. Ryan & Laura E. Simmons, Cornerstone Research, Securities Class Action Settlements: 2012 Review and Analysis 2 (2013), available at http://www.cornerstone.com/files/upload /Cornerstone_Research_2012_Settlements.pdf (on file with the Columbia Law Review) (reporting $2.9 billion in settlements in 2012 and $1.4 billion in 2011). For antitrust filing statistics, see Sourcebook of Criminal Justice Statistics Online, Table 5.41.2012, State Univ. of N.Y. at Albany (2012), http://www.albany.edu/sourcebook/csv/t5412012.csv (on file with the Columbia Law Review) (reporting average of roughly 795 private antitrust lawsuits between 2007 and 2012). 7. James B. Helmer, Jr., How Great Is Thy Bounty: Relator’s Share Calculations Pursuant to the False Claims Act, 68 U. Cin. L. Rev. 737, 744 (2000) (touting “spectacular results” of FCA Amendments Act of 1986); see also False Claims Act: Hearing Before the Subcomm. on the Constitution & Civil Justice of the H. Comm. on the Judiciary, 113th Cong. 61 (2014) (prepared statement of John E. Clark) (asserting qui tam relators and their counsel “expose fraud schemes otherwise unknown to the government” while also “augmenting the government’s resources” by performing “time-intensive tasks of screening cases, interviewing witnesses, analyzing and organizing available evidence, evaluating legal merit, preparing and filing complaints”); id. at 10 (prepared statement of Sen. Chuck Grassley) [hereinafter Grassley Statement] (noting $42 billion in recoveries and asserting “no other law in existence has been more effective in battling fraud than the False Claims Act has in the past 25 years”); Brief Amicus Curiae of AARP in Support of Respondent at 4–6, Schindler Elevator Corp. v. United States ex rel. Kirk, 131 S. Ct. 1885 (2011) (No. 10-188), 2011 WL 288888 (declaring “FCA is the single most effective tool in the fight against fraud perpetrated against the government” and crediting qui tam provisions with achieving large recoveries for federal treasury while also combating substandard healthcare); Pamela H. Bucy, Private Justice, 76 S. Cal. L. Rev. 1, 53 (2002) (describing FCA’s qui tam regime as “extraordinarily successful as a regulatory tool”); Barton H. Thompson, Jr., The Continuing Innovation of Citizen Enforcement, 2000 U. Ill. L. Rev. 185, 230–35 (surveying qui tam enforcement and declaring regime sufficiently successful to justify its export to environmental protection); Jill E. Fisch, Class Action Reform, Qui Tam, and the Role of the Plaintiff, Law & Contemp. Probs., Autumn 1997, at 167, 198–202 (touting qui tam’s success and advocating its export to securities and antitrust). 8. Such claims have issued from multiple quarters. See, e.g., United States ex rel. Stevens v. Vt. Agency of Natural Res., 162 F.3d 195, 222 (2d Cir. 1998) (noting large recoveries have “spawned the growth of a ‘qui tam bar’”); The False Claims Act Correction Act (S. 2041): Strengthening the Government’s Most Effective Tool Against Fraud for the 21st Century: Hearing Before the S. Comm. on the Judiciary, 110th Cong. 84 (2008) (statement of John T. Boese, Partner, Fried, Frank, Harris, Shriver, & Jacobson LLP) (criticizing “lottery-like aspect” of regime); Brief Amicus Curiae of the Chamber of Commerce of the United States of America in Support of Petitioner at 3, Schindler Elevator Corp., 131 S. Ct. 1885 (No. 10-188), 2010 WL 3452203 (“The FCA’s relaxed intent standard, harsh treble and per-claim damages specifications, and fee-shifting provision 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1917 driven the law in politically unaccountable directions, exploiting ambiguous regulatory mandates to assert claims that agencies and prosecutors are unwilling to embrace or tethering the FCA’s open-ended antifraud mandate to violations of separate statutes for which Congress did not provide a private right of action.9 On this view, qui tam litigation is not just socially inefficient. It is also democratically illegitimate. have combined to produce an expansive cottage industry of bounty-seeking relators.”); John Ashcroft et al., Whistleblowers Cash In, Unwary Corporations Pay, 40 Hofstra L. Rev. 367, 370 (2011) (“[Q]ui tam lawsuits represent a unique cottage industry within the legal sector . . ..”); Sean Elameto, Guarding the Guardians: Accountability in Qui Tam Litigation Under the Civil False Claims Act, 41 Pub. Cont. L.J. 813, 816 (2012) (noting “eruption” of qui tam suits); Michael Rich, Prosecutorial Indiscretion: Encouraging the Department of Justice to Rein in Out-of-Control Qui Tam Litigation Under the False Claims Act, 76 U. Cin. L. Rev. 1233, 1237 (2008) (noting “avalanche of qui tam suits”); Robert Stowe England, Hunting Bounty, Nat’l Rev., June 12, 1995, at 56, 56 (noting “flood of mostly minor suits”); False Claims Gold Rush, Wall St. J. (May 2, 2008, 12:01 AM), http://online.wsj.com/news/articles/SB120968865969361371 (on file with the Columbia Law Review) (describing FCA amendments as sparking “legal gold rush”); Amy Kolz, The Professional, Am. Law., June 1, 2010, available at LexisNexis (criticizing efforts of “professional” and “serial” relators who file multiple qui tam suits); Neil Weinberg, The Dark Side of Whistleblowing, Forbes (Mar. 14, 2005, 12:00 AM), http://www.forbes.com/ forbes/2005/0314/090.html (on file with the Columbia Law Review) (warning of relator- bar pursuit of “jackpot justice”). 9. See James F. Blumstein, What Precisely Is “Fraud” in the Health Care Industry?, Wall St. J., Dec. 8, 1997, at A25 (on file with the Columbia Law Review) (decrying tendency of “private whistleblowers” to bring kickback claims targeting practices that in fact improve healthcare delivery and so would not normally attract public prosecutors); see also Joan H. Krause, Regulating, Guiding, and Enforcing Health Care Fraud, 60 N.Y.U. Ann. Surv. Am. L. 241, 265 (2004) [hereinafter Krause, Regulating] (noting qui tam cases in healthcare context “may be used to achieve substantive results that neither Congress nor HHS has been willing to embrace”); Joan H. Krause, Twenty-Five Years of Health Law Through the Lens of the Civil False Claims Act, 19 Annals Health L. 13, 15 (2010) [hereinafter Krause, Twenty-Five] (“Relators and their attorneys have been instrumental in pushing the boundaries of the statute . . . [to other] laws that do not themselves permit a private right of action . . ..”); Dayna Bowen Matthew, An Economic Model to Analyze the Impact of False Claims Act Cases on Access to Healthcare for the Elderly, Disabled, Rural and Inner-City Poor, 27 Am. J.L. & Med. 439, 440–42, 447–50 (2001) (criticizing tethering of FCA to antikickback laws for which Congress did not provide private right of action); Lisa Michelle Phelps, Note, Calling Off the Bounty Hunters: Discrediting the Use of Alleged Anti-Kickback Violations to Support Civil False Claim Actions, 51 Vand. L. Rev. 1003, 1027–28, 1045 (1998) (arguing FCA’s use as “statutory vehicle” to enforce antikickback statutes is “in direct contradiction to the enforcement scheme Congress created”); James F. Blumstein, Rationalizing the Fraud and Abuse Statute, Health Aff., Winter 1996, at 118, 118–19 (distinguishing between FCA suits targeting “raw,” conventionally defined frauds and more arguable violations); Gregory T. Jaeger & Jonathan L. Diesenhaus, Fractious Fraud Fights, Legal Times, Oct. 21, 1996, at S32 (explaining increase in FCA cases has “spawned new theories of liability, many of which stretch the boundaries of the FCA beyond its logical jurisdictional limits”); Robert Salcido, The Government Unleashes the Stark Law to Enforce the False Claims Act: The Implications of the Government’s Theory for the Future of False Claims Act Enforcement, Health Law., Aug. 2001, at 1, 7 [hereinafter Salcido, Unleashes] (decrying use of FCA “by private individuals even though [federal self-referral law] itself does not permit any private right to action”). 1918 COLUMBIA LAW REVIEW [Vol. 114:1913

In some ways, the growing battle over qui tam exemplifies an undis- ciplined and often dysfunctional debate about litigation that has long afflicted American law and politics.10 But the above claims—particularly the notion that qui tam litigation has strayed beyond congressional purposes or is “undemocratic”—also raise profound questions about the use of private enforcement as a regulatory tool that extend well beyond the FCA, from securities and antitrust to civil rights and consumer protection. An initial set of questions concerns the ability of public actors, especially legislators, to manipulate the quantity of lawsuits in such regimes. Are litigation flows the rational, predictable result of structural shifts in the litigation environment, especially legislated incentives (e.g., available damages or attorneys’ fees)? Or do other factors, such as sociocultural trends, the vicissitudes of press coverage, or the organization of the plaintiffs’ bar, swamp direct microeconomic incentives, thus weaken- ing or even defeating public control? A growing empirical literature, much of it focused on job-discrimination litigation, has begun to explore such questions.11 Yet there has been only limited scholarly effort to address a related set of questions about changes over time in the character of lawsuits in litigation regimes and the subtly different political- and democratic- accountability concerns that result. In particular, how does private enforcement shape the elaboration of legal mandates over the life of a regulatory regime, especially compared to exclusively public enforcement by prosecutors or administrative agencies? Are “private attorneys general” immune, as qui tam’s critics complain, from meaningful public control where enforcement efforts stray beyond legislative purposes? In short, how, if at all, is making law with private lawsuits different from making law by other means? With rare exceptions, legal scholars have ducked these questions, falling back on a pair of abstract and conflicting claims: (i) that profit-motivated private enforcers supply salutary forms of legal innovation12 and (ii) that private enforcement and the legal innova-

10. See generally Thomas F. Burke, Lawyers, Lawsuits, and Legal Rights: The Battle over Litigation in American Society (2002) (critiquing American litigation politics). 11. See, e.g., Sean Farhang, Congressional Mobilization of Private Litigants: Evidence from the Civil Rights Act of 1991, 6 J. Empirical Legal Stud. 1, 27 (2009) [hereinafter Farhang, Mobilization] (finding increase in Title VII filings following Civil Rights Act of 1991’s provision of compensatory and punitive damages); Sean Farhang & Douglas M. Spencer, Legislating Incentives for Attorney Representation in Civil Rights Litigation, 2 J.L. & Cts. 241, 241–42 (2014) (finding increase in represented plaintiffs following Civil Rights Act of 1991). 12. See, e.g., James J. Park, Rules, Principles, and the Competition to Enforce the Securities Laws, 100 Calif. L. Rev. 115, 159 (2012) (noting entrepreneurial private enforcers “invest resources in pursuing innovative theories of wrongdoing”); Michael Selmi, Public vs. Private Enforcement of Civil Rights: The Case of Housing and Employment, 45 UCLA L. Rev. 1401, 1404 (1998) (arguing private plaintiffs, not government agencies, bring housing- and job discrimination suits asserting “cutting edge” claims). For examples of innovative enforcement by entrepreneurial counsel, see, e.g., Matthew C. Stephenson, Public Regulation of Private Enforcement: The Case for 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1919 tions it generates are dangerously immune from democratically accountable control.13 The result is a wide gap in scholarly understanding, both theoretical and empirical, of how, or to what extent, enforcement choices shape legal development over time.14 Closing this gap is important. One of the most significant developments in the American regulatory state in recent decades is a marked shift away from administrative regulation and enforcement and toward the use of private lawsuits to regulate social and economic behavior.15 Across a range of regulatory contexts, private, not public, enforcement

Expanding the Role of Administrative Agencies, 91 Va. L. Rev. 93, 112–13 (2005) [hereinafter Stephenson, Public Regulation] (describing how private counsel pioneered “supplemental enforcement projects” in environmental suits and litigated many leading antitrust precedents). 13. The strongest critic in this regard is Martin Redish, who has vigorously questioned the democratic legitimacy of aggregated litigation in particular. See Martin H. Redish, Wholesale Justice: Constitutional Democracy and the Problem of the Class Action Lawsuit 228 (2009) [hereinafter Redish, Wholesale Justice] (critiquing effect of class- action litigation on democracy); see also Richard A. Nagareda, Mass Torts in a World of Settlement 220 (2007) (arguing mass-tort lawsuits “involve governance, not litigation” and proposing reforms empowering administrative agencies in settlement efforts); Richard A. Nagareda, The Preexistence Principle and the Structure of the Class Action, 103 Colum. L. Rev. 149, 149–59 (2003) (noting “class action has emerged as a system of law reform that rivals conventional legislation” and proposing procedural reforms designed to achieve “proper allocation of lawmaking power between class counsel and courts on the one hand and democratically accountable institutions on the other”); Martin H. Redish, Class Actions and the Democratic Difficulty: Rethinking the Intersection of Private Litigation and Public Goals, 2003 U. Chi. Legal F. 71, 73 (arguing class actions have “undermined the foundational precepts of American democracy . . . by effectively transforming the essence of the governing substantive law that the class action has been created to enforce” without “democratic process of legislative amendment”); Stephenson, Public Regulation, supra note 12, at 119 (“As neither the citizens bringing private enforcement suits nor the judges who decide them are subject to electoral discipline, private enforcement may undermine a valuable democratic feature of American governance.”); Richard B. Stewart & Cass R. Sunstein, Public Programs and Private Rights, 95 Harv. L. Rev. 1193, 1292 (1982) (arguing private enforcement can “undermin[e] the advantages of political accountability . . . that administrative regulation was designed to provide”). 14. As discussed below, a vast literature—mostly in the law-and-economics vein— models the social-welfare and efficiency implications of the public–private enforcer choice. See infra note 19 and accompanying text. Interestingly, one of the seminal contributions to the law-review literature does not address private litigation’s effect on legal development at all, instead casting private enforcement’s principal virtue as its ability to help “stabili[ze]” legal norms via mechanical, profit-driven implementation. See John C. Coffee, Jr., Rescuing the Private Attorney General: Why the Model of the Lawyer as Bounty Hunter Is Not Working, 42 Md. L. Rev. 215, 226–30 (1983) [hereinafter Coffee, Rescuing]. 15. See Sean Farhang, The Litigation State: Public Regulation and Private Lawsuits in the United States 10–13 (2010) [hereinafter Farhang, Litigation State] (noting in past decade on average 165,000 lawsuits filed per year to enforce federal statutes “spann[ing] the waterfront of federal policy”). 1920 COLUMBIA LAW REVIEW [Vol. 114:1913 now dominates the field.16 But strong forces are also pushing in the opposite direction, including judicial constriction of class actions and punitive damages and administrative use of regulatory preemption, whereby federal agencies displace private rights of action entirely.17 These trends have brought renewed debate about the choice between public and private enforcement as regulatory tools and renewed interest in understanding the core attributes of each.18

16. See, e.g., Daniel A. Crane, Optimizing Private Antitrust Enforcement, 63 Vand. L. Rev. 675, 675–76 (2010) (noting private antitrust enforcement “vastly outstrips public enforcement”); David Freeman Engstrom, Agencies as Litigation Gatekeepers, 123 Yale L.J. 616, 689–90 (2013) [hereinafter Engstrom, Gatekeepers] (noting dominance of private enforcement in job-discrimination context); Park, supra note 12, at 159 (“Much securities enforcement, especially fraud enforcement, occurs through private class actions.”). 17. See, e.g., Robert H. Klonoff, The Decline of Class Actions, 90 Wash. U. L. Rev. 729, 745–823 (2013) (assessing limits on class actions); Catherine M. Sharkey, Punitive Damages as Societal Damages, 113 Yale L.J. 347, 375–80 (2003) (assessing limits on punitive damages). See generally Thomas O. McGarity, The Preemption War: When Federal Bureaucracies Trump Local Juries (2008) (providing overview of regulatory preemption); J. Maria Glover, The Structural Role of Private Enforcement Mechanisms in Public Law, 53 Wm. & Mary L. Rev. 1137 (2012) (describing broader trend against private enforcement). 18. For a sampling of scholarly contributions along these lines, see generally Farhang, Litigation State, supra note 15, at 10 (observing private litigants pursue “overwhelming majority of cases” implicating federal regulatory policy); Stephen B. Burbank, Sean Farhang & Herbert M. Kritzer, Private Enforcement, 17 Lewis & Clark L. Rev. 637, 640 (2013) (exploring “social and political character of private enforcement”); Crane, supra note 16 (discussing private antitrust enforcement); David Freeman Engstrom, Harnessing the Private Attorney General: Evidence from Qui Tam Litigation, 112 Colum. L. Rev. 1244 (2012) [hereinafter Engstrom, Harnessing] (conducting empirical analysis of qui tam lawsuits and drawing lessons for private enforcement generally); Myriam E. Gilles, Reinventing Structural Reform Litigation: Deputizing Private Citizens in the Enforcement of Civil Rights, 100 Colum. L. Rev. 1384, 1387–88 (2000) (suggesting model to “harness[] the power of private citizens to reform unconstitutional practices”); Margaret H. Lemos, Aggregate Litigation Goes Public: Representative Suits by State Attorneys General, 126 Harv. L. Rev. 486 (2012) (criticizing aggregate litigation undertaken by state attorneys general for a lack of procedures protecting citizens who disagree); Margaret H. Lemos & Max Minzner, For-Profit Public Enforcement, 127 Harv. L. Rev. 853, 895–903 (2014) (describing how financial incentives affect public enforcement); R. Preston McAfee, Hugo M. Mialon & Sue H. Mialon, Private v. Public Antitrust Enforcement: A Strategic Analysis, 92 J. Pub. Econ. 1863 (2008) (modeling private and public enforcement of antitrust laws and noting circumstances yielding socially optimal results); Park, supra note 12, at 178 (analyzing proposals for centralizing securities enforcement, then proposing “two-tier system of enforcement that might better utilize the comparative advantages of enforcers”); Amanda M. Rose, The Multienforcer Approach to Securities Fraud Deterrence: A Critical Analysis, 158 U. Pa. L. Rev. 2173, 2200–03 (2010) (discussing potential costs of private enforcement compared to “well-incentivized public enforcer”); Amanda M. Rose, Reforming Securities Litigation Reform: Restructuring the Relationship Between Public and Private Enforcement of Rule 10b-5, 108 Colum. L. Rev. 1301, 1330–49 (2008) (comparing relative effectiveness of public versus private enforcement for Rule 10b-5 class action by examining assumptions about public enforcement). For perspectives on regulatory choice, including the choice between agencies and courts and between public and private enforcement, see generally 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1921

This Article wades into this broader debate. Using qui tam’s recent and dramatic emergence, it develops and empirically tests competing accounts of “private enforcement’s pathways”—that is, the ways private enforcement efforts, acted upon by various features of the litigation environment, evolve over time and, in turn, shape the elaboration of legal mandates. The most basic aim is to focus sustained attention on the critical temporal dimension of private enforcement and the resulting calibration and control challenges that inhere in its choice as a regulatory tool. In so doing, the analysis offers policymakers a fuller set of tools with which to evaluate the legal and policy consequences of regulatory choices. More broadly, the analysis makes a novel contribution to the otherwise vast scholarly literature on the choice between public and private enforcement.19 In particular, mapping private enforcement’s pathways pushes past the standard scholarly focus on the relative efficiency of competing enforcement modes—that is, whether public or private enforcement delivers desired deterrence at lower social cost20— and instead explores how the choice of enforcer can shape the content of the law itself. The rest of this Article proceeds in five steps. Part I begins by sketching a pair of competing accounts of private enforcement’s pathways that have long lurked just below the surface in scholarly debate about the possibilities and limits of deputizing “private attorneys general” to enforce statutory mandates. The standard account sees profit-driven private enforcement as a mechanistic and manipulable engine of regulatory output that frictionlessly adapts to shifts in litigation incentives (such as available damages or attorneys’ fees) and also the amount of misconduct, thus picking up slack for regulatory failures elsewhere in the system. A more jaundiced account, by contrast, sees private enforcement as contingent, not mechanistic, and prone to chronic information failures that cause litigation activity to lurch from socially inefficient “explosions” to equally costly “droughts.” Profit-seeking private enforcers, runs this alternative view, will also relentlessly push into legal interstices, exploiting statutory and regulatory ambiguities in suits against much or all of an

Regulation Versus Litigation: Perspectives from Economics and Law (Daniel P. Kessler ed., 2011) (compiling theoretical and empirical works analyzing trade-off between regulation and litigation approaches). 19. See, e.g., A. Mitchell Polinsky & Steven Shavell, The Theory of Public Enforcement of Law, in 1 Handbook of Law and Economics 403, 420–22 (A. Mitchell Polinsky & Steven Shavell eds., 2007) (summarizing large scholarly literature on theory of public enforcement and comparing it to practice); Reza Rajabiun, Private Enforcement of Law, in Criminal Law and Economics 60, 60–64 (Nuno Garoupa ed., 2009) (same as to private enforcement). 20. See, e.g., Andrei Shleifer, The Failure of Judges and the Rise of Regulators 6 (2012) (“The choice between regulation, litigation, or a mixture of the two is then a choice of the efficient strategy of enforcement of socially desirable conduct.”); Polinsky & Shavell, supra note 19, at 406 (casting public–private choice in deterrence and social-cost terms); Rajabiun, supra note 19, at 61 (same). 1922 COLUMBIA LAW REVIEW [Vol. 114:1913 industry rather than targeting the patently illegal conduct of a few malefactors. Growing scale and scope are worrying, the rival account concludes, because courts, legislators, and agencies can only imperfectly police private enforcement efforts, and also because litigation can yield path dependencies where, for instance, defendants who lose early lawsuits lobby political actors to clarify the law to ensure their competitors are subject to the same liability. Private enforcement can thus drive law down interpretive pathways it would not travel if enforcement were in purely public hands.21 Parts II through IV then use qui tam litigation brought under the FCA as an empirical laboratory to adjudicate between these competing accounts. Part II motivates the empirical analysis by offering a brief overview of the regime’s core workings, making the case for qui tam’s use as a case study, and reviewing a litany of anecdotal claims made about qui tam’s pathways in Supreme Court amicus filings, congressional testimony, and articles in the popular and academic press. Part III turns to empirics and uses a novel data set of nearly 6,000 qui tam lawsuits filed since 1986 to test claims about qui tam’s recent growth. The findings show an unmistakable rise in qui tam enforcement activity. But along numerous dimensions—from filing counts to more complicated measures, such as per-filing recovery amounts—the findings reject the notion that recent growth in qui tam litigation constitutes a “gold rush” of enforcement activity or is otherwise out of control.22 As research across a range of other contexts shows, litigation “explosions” are often less explosive than critics claim.23 Qui tam appears to be no exception. Part IV moves to subtler questions about whether qui tam enforcement efforts have changed in their essential character and, further, whether those changes have driven the elaboration of the FCA’s open-ended antifraud mandate in ways that have defied political control. Though these issues are less susceptible to direct empirical testing than Part III’s data-driven accounting of qui tam’s recent growth, the findings are also more concerning. For instance, the data show a pronounced rise in the number of defendants per qui tam action.24 This lends credence to

21. See infra Part I (discussing theoretical framework of private enforcement). 22. See infra Part III.A (evaluating filing counts); infra Part III.B (analyzing recovery amounts); infra Part III.C (analyzing per-filing recovery amounts). 23. See Arthur R. Miller, The Pretrial Rush to Judgment: Are the “Litigation Explosion,” “Liability Crisis,” and Efficiency Clichés Eroding Our Day in Court and Jury Trial Commitments?, 78 N.Y.U. L. Rev. 982, 985–96 (2003) (“[I]n spite of these phenomena, there is contrary evidence indicating that the claims of the alleged ‘litigation explosion’ are exaggerated; indeed, that evidence casts doubt on the very existence of a significant increase.”); see also Marc Galanter, The Day After the Litigation Explosion, 46 Md. L. Rev. 3, 37–39 (1986) (noting litigation “explosion” results from “series of local changes” and does not imply “heightened appetite for adversarial combat”). 24. See infra Figure 7 (plotting logarithm of number of defendants in qui tam actions filed since 1986 and fitting regression model). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1923 anecdotal claims that qui tam litigation has steadily expanded in its scale and regulatory scope, moving away over time from targeting uncontro- versial frauds, such as patently illegal overbilling for Medicare services, and instead moving towards exploiting regulatory ambiguities in suits targeting more arguable wrongdoing engaged in by much or all of an industry. The qualitative case studies of several marquee flavors of qui tam litigation that follow enrich these findings, offering substantial evidence that qui tam privateers have pushed into regulatory gaps left by legislative and administrative inertia.25 These same enforcement efforts have also targeted conduct that, far from eluding the notice or imagina- tion of legislators or regulators, was in fact well known to the government.26 To be sure, one might question the significance of such findings regarding the overall degree of political control within the system. Given the counterfactual nature of much of the inquiry, reasonable minds might disagree about whether the profiled qui tam actions proceeded with the implicit blessing of Congress or agencies with regulatory or enforcement authority—and thus whether the legal evolution of the FCA’s antifraud mandate would have looked materially different had enforcement remained solely in public hands. The case studies also raise complex questions about whether qui tam enforcement efforts can be cast as playing a salutary agency-forcing or anticapture role. Finally, the DOJ’s powerful oversight authority within the FCA regime—including its statutory authority to intervene in qui tam suits and take over their control or dismiss them outright—raises the possibility that the FCA should not be considered a private enforcement regime at all, but rather a complex public–private hybrid or even a mostly public one. Yet whatever one’s view on this latter point, the evidence further suggests that DOJ control over qui tam litigation may be loosening as the regime shifts into a higher-volume retail mode.27 This is troubling because the decoupling of qui tam enforcement from DOJ control may be occurring just as the regime enters a period of significant further growth and innovation. Part V brings together the theory and findings and outlines their implications for renewed scholarly debate about the choice between public and private enforcement across a wide range of regulatory contexts and a broader, but related, debate about the complex relationship between litigation and democracy. A more rigorous theoretical and empirical understanding of private enforcement’s pathways, the begin- nings of which are offered here, is essential to judging the merits and

25. See infra Part IV.B (using case studies to show tendency of private enforcement to exploit regulatory ambiguities left by legislative and administrative inertia). 26. See infra Part IV.B (discussing case studies where qui tam litigation targeted well- known conduct). 27. See infra Part IV.C (discussing empirical evidence of DOJ’s shifting qui tam oversight role). 1924 COLUMBIA LAW REVIEW [Vol. 114:1913 demerits of private enforcement of public law, and its choice over regulatory alternatives, in the qui tam context and beyond.

I. PRIVATE ENFORCEMENT’S PATHWAYS:THEORETICAL FRAMEWORK

What drives the volume and nature of lawsuits in litigation regimes? That question is of enormous practical importance to the optimal design of regulatory regimes that deploy private litigation as an enforcement tool—from securities and antitrust to civil rights and consumer protection. Indeed, in these and other contexts, the legitimacy of choosing private enforcement over regulatory alternatives, such as enforcement by prosecutors or administrative agencies, turns on at least some measure of public control over the amount and character of private lawsuits. If, by contrast, “private attorneys general” are immune from public control, or if forces beyond the legislative design substantially drive the quantity or quality of litigation efforts, one might be less willing to put the reins in private hands. Yet few studies offer sustained theoretical analysis of the forces that shape private litigation efforts over time or the degree to which such efforts are subject to meaningful political control. This Part begins to remedy that situation by building upon longstanding scholarly debate about the choice between public and private enforcement and sketching a pair of competing accounts of what was earlier dubbed “private enforcement’s pathways.” This approach sheds new theoretical light on the calibration and control challenges legislators face in deploying private enforcement as a regulatory tool. It also yields some testable hypotheses in advance of the empirical study of qui tam litigation to come in Parts II through IV.

A. Is Private Enforcement Rational? Litigation Quantity and the Calibration Challenge The standard law-and-economics view of litigation behavior proceeds from the assumption that plaintiffs will sue whenever the expected value of doing so exceeds expected cost.28 Tethered to the profit motive, private enforcement will thus operate in a kind of market-clearing, “autopilot” mode.29 Assumptions about the essential profit motivation of private enforcers offer a useful starting point for thinking about the behavior of private litigation regimes by highlighting some of private enforcement’s benefits and costs compared to regulatory alternatives. For instance, private

28. See Farhang, Litigation State, supra note 15, at 22 (summarizing standard model of decision to litigate as plaintiff calculating probability of prevailing times expected value, minus costs of filing suit). 29. See id. at 20 (using “autopilot” locution); cf. Rajabiun, supra note 19, at 64–65 (likening private enforcement regimes to “price system”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1925 enforcement’s champions assert that the profit-motivated nature of private enforcement renders it more cost-effective and nimbler than poorly incentivized or sclerotic public enforcement bureaucracies.30 A further claim is that profit-seeking private enforcers will provide a politically insulated, “failsafe” source of enforcement if public enforcement agencies, such as the Securities and Exchange Commission or the Federal Trade Commission, suffer from lassitude, resource constraints, or “capture” by regulated interests.31 Critics, on the other hand, invoke private enforcement’s “autopilot” nature to highlight its potential inefficiencies. A standard critique holds that profit-driven private enforcers will overenforce by suing whenever it pays to do so, even where the social costs incurred—e.g., total litigation and other transaction costs, or the costs imposed on affected communities—exceed any benefit.32 Private enforcement’s decentralized and

30. See, e.g., Stewart & Sunstein, supra note 13, at 1298 (noting public agencies suffer from “diseconomies of scale, given multiple layers of decision and review”). 31. See Stephenson, Public Regulation, supra note 12, at 110 (“[A] potential benefit of private enforcement suits is that they can correct for agency slack—that is, the tendency of government regulators to underenforce certain statutory requirements because of political pressure, lobbying by regulated entities, or the laziness or self-interest of the regulators themselves.”); see also Matthew D. Zinn, Policing Environmental Regulatory Enforcement: Cooperation, Capture, and Citizen Suits, 21 Stan. Envtl. L.J. 81, 132 (2002) (noting possible anticapture role of “citizen suits” brought under federal environmental law). For current scholarly thinking on the capture concept, see generally Preventing Regulatory Capture: Special Interest Influence and How to Limit It 11 (Daniel Carpenter & David A. Moss eds., 2014) [hereinafter Preventing Regulatory Capture] (analyzing special interest influence in administrative policymaking, existing mechanisms for limiting such influence, and circumstances under which such mechanisms may succeed); David Freeman Engstrom, Corralling Capture, 36 Harv. J.L. & Pub. Pol’y 31, 31–39 (2013) (asserting significance of definitions, contextualization, and comparison in discussing capture). 32. See Steven Shavell, The Fundamental Divergence Between the Private and the Social Motive to Use the Legal System, 26 J. Legal Stud. 575, 578 (1997) (“[W]hen a person brings suit, he bears only his own legal expenses; he does not take into account that his suit will cause the defendant and possibly the court to incur legal expenses as well; a bias toward excessive suit is thus engendered.”); see also Eugene Bardach & Robert A. Kagan, Going by the Book: The Problem of Regulatory Unreasonableness 25 (1982) (noting related concern that mechanistic private enforcement may apply generic legal rules to factual settings in ways that do not align with statutory purposes, causing enforcement to exhibit “site-level unreasonableness”); Stephenson, Public Regulation, supra note 12, at 115 (“[P]rivate parties may be less sensitive than government agencies to the economic and social costs of particular enforcement actions, such as the disruptive impact on affected communities . . . .”). Even apart from this dynamic, the standard law- and-economics analysis holds that private enforcement will be inefficient relative to public enforcement because prosecutors can achieve desired deterrence at lower social cost by imposing maximal sanctions on relatively few defendants and exercise discretionary nonenforcement as to the rest. By contrast, increasing the damages available to private enforcers merely yields ever-greater private investment in enforcement. See William M. Landes & Richard A. Posner, The Private Enforcement of Law, 4 J. Legal Stud. 1, 15 (1975) (offering seminal analysis). But see Engstrom, Gatekeepers, supra note 16, at 631 & n.40 (noting limitations of this analysis). 1926 COLUMBIA LAW REVIEW [Vol. 114:1913 unyielding nature likewise exacts an efficiency toll by inviting duplicative enforcement actions in which private enforcers “piggyback” on public enforcers and one another33 and by disrupting the subtle cooperative relationships that arise between regulators and regulatory targets and ensure a degree of regulatory “coherence.”34 Yet if simple assumptions about profit motivation help highlight the costs and benefits of private enforcement compared to regulatory alternatives, then the debate has also tended to gloss over deeper, and more fundamental, questions about the degree to which private enforcement is subject to meaningful calibration and control. For instance, private enforcement’s champions, and even many of its critics, seem to take for granted a measure of public control over litigation flows. Indeed, if private enforcers tend to operate on “autopilot,” then a legislature that desires relatively more or less enforcement need only dial up or down payouts (e.g., multiple or punitive damages, or damages caps) or attorneys’ fees. The result is a mostly optimistic view of private enforcement as a mechanistic and manipulable—if at times inefficient— substitute for public enforcement.35 Lurking beneath existing scholarly accounts, however, is a competing, and less optimistic, view. As an initial matter, some dismiss the significance of microeconomic incentives altogether, focusing instead on the role of extralegal factors, such as broad-scale sociocultural forces,36 macroeconomic health,37 or noneconomic litigant motives,38 in deter-

33. See Stephenson, Public Regulation, supra note 12, at 128 & n.117 (noting piggybacking concern); see also Coffee, Rescuing, supra note 14, at 222 (noting “recurring pattern . . . under which the private attorney general simply piggybacks on the efforts of public agencies—such as the SEC, the FTC, and the Antitrust Division of the Department of Justice—in order to reap the gains from the investigative work undertaken by these agencies”). 34. See Richard L. Revesz, Specialized Courts and the Administrative Lawmaking System, 138 U. Pa. L. Rev. 1111, 1117 (1990) (defining regulatory “coherence” as statutory scheme “reflect[ing] a unitary vision” rather than mere absence of contradictory rules); Stewart & Sunstein, supra note 13, at 1227 & n.134 (noting subtle cooperative relationships often emerge between regulators and regulatory targets). 35. See, e.g., A. Mitchell Polinsky, Private Versus Public Enforcement of Fines, 9 J. Legal Stud. 105, 108 (1980) (concluding private enforcement levels can, if payouts are set properly, be made to approximate social ideal in competitive markets); see also A. Mitchell Polinsky & Yeon-Koo Che, Decoupling Liability: Optimal Incentives for Care and Litigation, 22 RAND J. Econ. 562, 564–68 (1991) (showing how decoupling liability and bounty can optimize litigation levels at lowest social cost). 36. See, e.g., Lawrence M. Friedman, Total Justice 5 (1994) (assessing role of “legal culture” in shaping litigation activity); Bayless Manning, Hyperlexis: Our National Disease, 71 Nw. U. L. Rev. 767, 767 (1977) (suggesting strong cultural role in litigation rates). 37. See Tonja Jacobi, The Role of Politics and Economics in Explaining Variation in Litigation Rates in the U.S. States, 38 J. Legal Stud. 205, 214 (2009) (offering theoretical and empirical tests of notion that litigation increases in economic growth). 38. See Farhang, Mobilization, supra note 11, at 7 (reviewing literature on psychological and other nonpecuniary motivations to sue); see also Theodore Eisenberg & Henry S. Farber, The Litigious Plaintiff Hypothesis: Case Selection and Resolution, 28 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1927 mining litigation flows. On this view, the volume of lawsuits will turn less on core structural features of a litigation regime or even the microeconomic incentives facing plaintiffs and their counsel, and more on other, potentially random aspects of the litigation environment.39 More fundamentally, and even if private enforcement is at least modestly responsive to legislative incentives, simple assumptions about private enforcement’s “autopilot” nature yield a radically incomplete theory of litigation behavior and legislative control. One problem is that much of the theorizing that extends from assumptions about private profit motivation tends to take highly static forms. But in reality, litigant pursuit of profit is likely to work in tandem with dynamic aspects of the litigation environment, such as the structure of the plaintiffs’ bar, in shaping litigation flows.40 When a legislature mints a private right of action, it induces private actors, particularly plaintiff-side law firms, to invest in enforcement capacity, such as regime-specific expertise or a firm-level infrastructure for identifying clients.41 These investments may shape litigation flows over time. For instance, one might expect a net decrease in litigation levels over time as specialized plaintiff-side firms move to the top of referral networks and begin to serve a salutary

RAND J. Econ. S92, S92 n.1 (1997) (acknowledging “rich literature covering suit, settlement, and trial of civil cases” but lack of such literature regarding impact of plaintiff filing process on subsequent events and outcomes); Robert J. MacCoun, Voice, Control, and Belonging: The Double-Edged Sword of Procedural Fairness, 1 Ann. Rev. L. & Soc. Sci. 171, 171–72 (2005) (discussing plaintiff interest in “procedural justice” and reviewing literature on noneconomic motivations to sue). For a discussion of litigation efforts as part of a broader political-advocacy strategy, see Michael McCann, Rights at Work: Pay Equity Reform and the Politics of Legal Mobilization 5–6 (1994). 39. It is now standard among social scientists to consider the “structural” factors— including sociological, economic, and political inputs—that determine litigation flows. See, e.g., David S. Clark, Civil Litigation Trends in Europe and Latin America Since 1945: The Advantage of Intracountry Comparisons, 24 Law & Soc’y Rev. 549, 550 (1990) (modeling litigation volume largely as function of extralegal factors); Jacobi, supra note 37, at 205 (same); Anne Morrison Piehl & Margo Schlanger, Determinants of Civil Rights Filings in Federal District Court by Jail and Prison Inmates, 1 J. Empirical Legal Stud. 79, 83–86 (2004) (modeling § 1983 filings as function of prison population). 40. See, e.g., John C. Coffee, Jr., Understanding the Plaintiff’s Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions, 86 Colum. L. Rev. 669, 706–11 (1986) [hereinafter Coffee, Understanding] (discussing factors influencing size of modern plaintiffs’ attorneys’ firms); Stephen C. Yeazell, Re-Financing Civil Litigation, 51 DePaul L. Rev. 183, 186 (2001) (offering classic account of role of evolving structure of plaintiffs’ bar in shaping legal contestation). 41. See Engstrom, Harnessing, supra note 18, at 1256–57 (discussing responses of private actors to narrowed standing); see also Robert L. Nelson & David M. Trubek, New Problems and New Paradigms in Studies of the Legal Profession, in Lawyers’ Ideals/ Lawyers’ Practices: Transformations in the American Legal Profession 1, 6 (Robert L. Nelson et al. eds., 1992) (“Perhaps what is most distinctive about American lawyers is the zeal with which they have developed new organizational forms for capturing particular segments of the market for legal services or for capitalizing on the economic opportunities presented by particular legal problems.”). 1928 COLUMBIA LAW REVIEW [Vol. 114:1913 screening function.42 It is just as possible, however, that specialization will increase litigation volume. Repeat plaintiffs’ counsel may take advantage of the lower marginal cost of litigating cases that specialization brings or develop abusive litigation tactics, using the threat of costly discovery to extract settlements in meritless cases or bringing “blunderbuss” suits against multiple defendants to obtain discovery that may reveal a winning case rather than performing careful prefiling inquiry.43 While it is hard to know which of these dynamics will prove more substantial, the broader point is that private enforcement’s regulatory output is likely to vary over time, as the regime reaches more mature stages of development, in ways that may be relevant to the choice of private litigation over regulatory alternatives. Taking a more dynamic, temporal view points to other critical ways private enforcement may systematically differ from regulatory alternatives: Compared to centralized public enforcers, decentralized private litigants may also be unable to accurately assimilate information about their probability of success.44 The result may be overentry or even a “gold rush” of lawsuits as decentralizated private enforcers, lacking centralized information gathering and so left to parse distorted press reports and unrepresentative published judicial decisions, pour resources into losing enforcement efforts.45 It is just as possible that decentralized, profit-

42. See Stephen J. Spurr, Referral Practices Among Lawyers: A Theoretical and Empirical Analysis, 13 Law & Soc. Inquiry 87, 100–02 (1988) (describing referral networks and their claim-sorting role); see also Sara Parikh & Bryant Garth, Philip Corboy and the Construction of the Plaintiffs’ Personal Injury Bar, 30 Law & Soc. Inquiry 269, 281 (2005) (discussing referral networks and their interactions with plaintiffs’ bar). For a description of the case-screening practices of the contingent-fee plaintiffs’ bar, see Herbert M. Kritzer, Risks, Reputations, and Rewards: Contingency Fee Legal Practice in the United States 67, 67–76 (2004). 43. See Engstrom, Harnessing, supra note 18, at 1259–63 (discussing disadvantages of specialization); see also Lucian Arye Bebchuk, Suing Solely to Extract a Settlement Offer, 17 J. Legal Stud. 437, 440–41 (1988) (offering general theory of litigation costs and settlement behavior). For concrete examples of pathologies that can accompany firm-level specialization, see Janet Cooper Alexander, Do the Merits Matter? A Study of Settlement in Securities Class Actions, 43 Stan. L. Rev. 497, 505–21 (1991) (describing “strike” suits in securities litigation); John C. Coffee, Jr., Class Wars: The Dilemma of the Mass Tort Class Action, 95 Colum. L. Rev. 1343, 1373–76 (1995) (discussing collusion among class attorneys in mass-tort litigation made possible by repeat play); James L. Stengel, The Asbestos End-Game, 62 N.Y.U. Ann. Surv. Am. L. 223, 234–36 (2006) (noting collusion between specialized plaintiff-side asbestos firms and unscrupulous physicians to manufacture fraudulent diagnoses and thus grow claim inventories). 44. See Glover, supra note 17, at 1180 (noting conditions under which public enforcers enjoy informational advantages, including, among others, where large amount of data is necessary to illuminate wrongdoing or where gauging liability requires comparative analysis across regulatory targets); see also Steven Shavell, Liability for Harm Versus Regulation of Safety, 13 J. Legal Stud. 357, 359, 369 (1984) (noting instances in which regulatory agency “may enjoy a positive [informational] advantage relative to private parties”). 45. See William Haltom & Michael McCann, Distorting the Law: Politics, Media, and the Litigation Crisis 159–74 (2004) (“[P]roliferation of . . . dramatized stories that 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1929 motivated private enforcers will not produce enough enforcement effort as flawed litigant perceptions prevent filings that would otherwise be socially, and even privately, beneficial.

FIGURE 1: EQUILIBRATION COSTS IN PRIVATE ENFORCEMENT REGIMES

Figure 1 makes these dynamics more concrete. It does so by plotting the hypothetical response of private enforcers to legislated shifts in litigation incentives (for instance, a legislative enactment relaxing statutory standing requirements in order to attract more plaintiffs and increase enforcement levels, or a contrary legislative enactment raising the substantive liability standard in order to decrease enforcement levels) and distorted information (for instance, media publicity recounting several idiosyncratically large litigation wins) alongside the socially optimal, and legislatively desired, quantity of enforcement. The result, as denoted by the shaded regions above and below the curves, is that decentralized private enforcers will chronically under- or overreact to shifts in litigation supported perceptions of rapidly increasing rates of litigation hindered rather than helped readers to understand the exceptional legal, political, and social developments at stake in mass tort claims.”). That plaintiffs uniquely suffer from information deficiencies is a standard claim among legal empiricists. See Daniel Kessler, Thomas Meites & Geoffrey Miller, Explaining Deviations from the Fifty-Percent Rule: A Multimodal Approach to the Selection of Cases for Litigation, 25 J. Legal Stud. 233, 243 (1996) (“Plaintiffs, in contrast, are unsophisticated and are unable to sort cases accurately into good or bad cases.”); George Loewenstein et al., Self-Serving Assessments of Fairness and Pretrial Bargaining, 22 J. Legal Stud. 135, 145–53 (1993) (reporting experimental results suggesting plaintiffs suffer from self-serving biases related to litigation outcomes). On media distortion, see Haltom & McCann, supra, at 159–74 (discussing skewed media coverage of lawsuits, particularly mass-tort cases); Robert J. MacCoun, Media Reporting of Jury Verdicts: Is the Tail (of the Distribution) Wagging the Dog?, 55 DePaul L. Rev. 539, 558–61 (2006) (analyzing media distortion in context of tort reform). 1930 COLUMBIA LAW REVIEW [Vol. 114:1913 incentives and new information relative to the social ideal, thereby imposing what one might call “equilibration costs” within the system. Of course, Figure 1 offers a highly stylized account, particularly in its suggestion that centralized public enforcement will, unlike private enforcement, approximate the social ideal. As a wave of recent scholarship suggests, public prosecutors and agency administrators have their own mix of politically inflected and careerist incentives that can cause public enforcement to deviate from the ideal.46 Moreover, and as previously suggested, sclerotic public bureaucracies may be less organization- ally dexterous than their private counterparts and thus less efficient at mobilizing or demobilizing enforcement capacity in response to shifts in litigation incentives or the litigation environment.47 Flat-footed public bureaucracies may thus impose their own, capacity-based type of equilibration cost that cuts against the centralized informational advantages they otherwise enjoy. More formal theoretical work could tease out these and other possibilities in building a more comprehensive framework for comparing competing public and private enforcement modes, a point addressed more fully in Part IV.48 Still, sketching private enforcement’s pathways, and thus considering private litigation flows through Figure 1’s temporal lens, permits a more complex, and much gloomier, set of predictions about private enforcement behavior that goes well beyond the usual story about the divergent private and social incentives to litigate.49 Indeed, chronic information failures may yield equilibration costs that rival or even exceed the costs that accrue from profit-driven overzealousness. In the extreme case,

46. See Lemos & Minzner, supra note 18, at 864–86 (noting ways in which public enforcement priorities will be politically determined); Jonathan R. Macey & Geoffrey P. Miller, Reflections on Professional Responsibility in a Regulatory State, 63 Geo. Wash. L. Rev. 1105, 1106, 1115–16 (1995) (noting “career-building” proclivities of government lawyers); McAfee, Mialon & Mialon, supra note 18, at 1872 (“[S]ome government actors are likely to be partly motivated by factors other than efficiency, including career concerns . . . .”). 47. See supra note 30 and accompanying text (noting cost and efficiency advantages of private enforcement over public enforcement). 48. A more comprehensive formal theory could also relax the assumption that public enforcement is monopolistic and instead take account of the fact, as just one example, that the federal prosecutorial apparatus is made up of regional U.S. Attorneys’ offices subject to only weak central control by “Main Justice” in Washington, D.C. See generally James Eisenstein, The U.S. Attorney Firings of 2006: Main Justice’s Centralization Efforts in Historical Context, 31 Seattle U. L. Rev. 219 (2008) (offering historical overview of conflict arising from Main Justice’s efforts to establish “headquarters control” and U.S. Attorneys’ efforts to maintain “field autonomy”). A further consideration in any formal modeling effort is that the social costs may not be large where the patterns of under- and overenforcement are symmetrically distributed around the ideal and those patterns are close enough in time. Where this is true, the costs of under- and overenforcement may simply even out, thus approximating the stylized public enforcement ideal. 49. See supra note 32 and accompanying text (cataloguing literature in support of theory private enforcers will disregard social costs of enforcement, leading to suboptimal overenforcement). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1931 these failures will cause private enforcement to lurch from socially inefficient “explosions” to equally costly “droughts” in ways that policymakers can neither predict nor control. Particularly in litigation regimes experiencing frequent changes in litigation incentives or the wider litigation environment, private enforcement may not just fall short of a mechanistic and dependable “failsafe.” It may also prove immune to meaningful calibration and control by legislative principals at all.50

B. Private Legal Innovation, Statutory “Drift,” and the Puzzle of Public Control Existing theories thus offer strikingly divergent predictions about private litigation flows and their manipulability by public policymakers. But litigation volume is not the only relevant measure of regulatory output or the only concern of policymakers. In order to chart private enforcement’s pathways, one must understand not just what drives the quantity of private lawsuits over time, but also the forces that drive changes in their nature and character.51 The effect private litigants have on the nature and character of enforcement efforts has long figured in scholarly debate about optimal institutional design. A common claim is that entrepreneurial private enforcers serve as incubators, bringing fresh ideas to the regulatory marketplace.52 Thus, Barton Thompson has shown how environmental groups have served as engines of innovation by developing novel remedial mechanisms in “citizen suits” brought under federal environmental law.53 Privately driven innovation can also take legal interpretive forms as profit-minded private enforcers develop and deploy new theories of liability under existing legal mandates and constraints.54 A good example

50. Cf. James G. March & Johan P. Olsen, Rediscovering Institutions: The Organizational Basis of Politics 55 (1989) (noting possibility that slow adaptation may leave political institutions and rules in a state of perpetual lag with their environment). 51. See, e.g., Park, supra note 12, at 129 (“Enforcement obviously differs in terms of quality, not just quantity. Thinking of enforcement as an output, much like a product, has the appeal of simplifying the analysis, but quantitative measures of enforcement do not capture the significant qualitative differences between enforcement cases.”); see also Nora Freeman Engstrom, Re-Re-Financing Civil Litigation: How Lawyer Lending Might Remake the American Litigation Landscape, Again, 61 UCLA L. Rev. Disc. 110, 112 (2013) (suggesting different types of financing alter litigation quantity and “character”); George L. Priest, Regulating the Content and Volume of Litigation: An Economic Analysis, 1 Sup. Ct. Econ. Rev. 163, 163–65 (1982) (considering use of litigation incentives, from fee shifts to settlement coercion, to shape not just litigation quantity but also litigation quality). 52. See Glover, supra note 17, at 1155–56 (discussing entrepreneurial benefits of private enforcement). 53. See Thompson, supra note 7, at 206–09 (describing private development of “supplemental environmental projects” as remedial mechanisms); see also William B. Rubenstein, On What a “Private Attorney General” Is—and Why It Matters, 57 Vand. L. Rev. 2129, 2160 n.112 (2004) (noting origins of cy pres doctrine in environmental citizen suit and class actions more generally). 54. See supra note 12 (collecting studies touting private enforcers’ production of “legal innovations”); see also Bucy, supra note 7, at 64 (noting private enforcers “have 1932 COLUMBIA LAW REVIEW [Vol. 114:1913 is the successful campaign by private lawyers (and law professors) in the 1970s to develop a theory of sexual harassment as cognizable sex discrimination under Title VII’s open-ended antidiscrimination mandate.55 However, there remain questions about the normative desirability of this process of privately driven legal innovation that have gone under- developed in the scholarly literature.56 A possible optimistic view parallels arguments lauding private enforcement as a salutary “failsafe.” For instance, the sexual-harassment example just noted might be thought to support the view that sophisticated entrepreneurial enforcers (whether plaintiffs or their counsel) can play a valuable regulatory role, updating statutes to confront social and economic problems in ways that a scle-

every incentive to push theories of liability as far as possible”); Stephen Daniels & Joanne Martin, “We Live on the Edge of Extinction All the Time”: Entrepreneurs, Innovation and the Plaintiffs’ Bar in the Wake of Tort Reform, in Legal Professions: Work, Structure, and Organization 149, 152 (Jerry Van Hoy ed., 2001) (analyzing “legal entrepreneurialism” of plaintiffs’ bar); Park, supra note 12, at 160 (noting tendency of plaintiff-side securities bar to bring “actions that have pushed the boundaries of the law”); Rubenstein, supra note 53, at 2152 (noting private enforcement would generate more prosecutorial innovation than “monopolistic government enforcer would produce”). 55. See Reva B. Siegel, A Short History of Sexual Harassment, in Directions in Sexual Harassment Law 1, 8–12 (Catharine A. MacKinnon & Reva B. Siegel eds., 2004) (tracing history of protests, scholarship, and litigant and judicial response to sexual harassment); see also Meritor Savings Bank v. Vinson, 477 U.S. 57, 63–69 (1986) (recounting campaign leading to EEOC and judicial sanction of sexual harassment as actionable sex discrimination). 56. As noted previously, a venerable literature considers the welfare and efficiency implications of the choice between public and private enforcement. See supra note 20 and accompanying text. But few scholarly studies consider the effect of delegation choices on legal development. For instance, both Lemos and Stephenson focus on how courts and agencies might differ in their abilities to deliver consistent application of legal principles across policy contexts and over time. See Margaret H. Lemos, The Consequences of Congress’s Choice of Delegate: Judicial and Agency Interpretations of Title VII, 63 Vand. L. Rev. 363, 364 (2010) (discussing debate over propriety and desirability of Congress’s delegation of lawmaking authority to agencies); Matthew C. Stephenson, Legislative Allocation of Delegated Power: Uncertainty, Risk, and the Choice Between Agencies and Courts, 119 Harv. L. Rev. 1035, 1038 (2006) (arguing “interpretive consistency” may influence congressional choice of delegation to agencies and courts). But both studies focus on the choice of adjudicator and differences in the decisionmaking attributes of courts and agencies, not the choice of enforcer or the ways in which the presence or absence of profit-motivated private enforcers shapes enforcement efforts and, thus, legal development. Other studies consider how private enforcement systematically differs from public enforcement but do not consider the effect on legal development. See, e.g., Peter Huber, Safety and the Second Best: The Hazards of Public Risk Management in the Courts, 85 Colum. L. Rev. 277, 277–78, 331–32 (1985) (arguing profit-motivated private enforcers will, relative to public regulators, inefficiently target novel “public” risks emanating from large entities developing new technologies that are generally safer than older background “private risks”); Park, supra note 12, at 119, 159–62 (distinguishing between enforcement efforts targeting clear “rules” and more open-ended, standard-like “principles,” and suggesting entrepreneurial private enforcers might be more inclined and better able to enforce the former). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1933 rotic, risk-averse, or “captured” public bureaucracy might not.57 Nor must this mean that private enforcers will simply be left to their own devices. After all, a trio of institutional actors—courts with adjudicatory authority, agencies with rulemaking or other oversight powers, and the legislature itself—stand ready to check or override private litigation efforts grounded in patently unfaithful statutory applications.58 More fundamentally, wise legislators may in fact prefer to delegate enforcement authority to private litigants (and, by extension, to courts) over delegation to public prosecutors or administrative agencies, particularly where these public actors are controlled by partisan opponents or are vulnerable to capture by well-organized interest groups.59 From this patchwork of theoretical possibilities emerges a complex but ultimately optimistic view in which private enforcers play a valuable innovative role, renovating obsolete statutes and perhaps even nudging agencies to do the same, yet remaining fully subject to public control where enforcement efforts stray too far beyond legislative purposes.60 Another less optimistic view proceeds by isolating the ways private enforcement is likely to shape, and be shaped by, the litigation environment over time. For instance, pursuit of profit will drive private legal innovation as entrepreneurial litigants and their counsel seek out new litigation opportunities.61 But that process will also, as just noted in Part II.A’s consideration of litigation flows, be molded by the dynamic evolution of the plaintiffs’ bar.62 As plaintiff-side firms working within a new litigation regime acquire expertise and accumulate capital, they will be

57. See supra notes 30–31 and accompanying text (reviewing literature making these claims). 58. The logic here resembles standard separation-of-powers models in which the House, Senate, and President each occupy a veto point but can coordinate to keep agency action within certain bounds. See, e.g., William N. Eskridge, Jr. & John Ferejohn, The Article I, Section 7 Game, 80 Geo. L.J. 523, 535–38 (1992) (modeling these shifts in statutory policy based on House, Senate, and presidential preferences). 59. See supra note 56 and accompanying text (reviewing studies of judicial delegations); see also Sean Farhang, Public Regulation and Private Lawsuits in the American Separation of Powers System, 52 Am. J. Pol. Sci. 821, 823–28 (2008) (showing congressional enactment of litigation-enhancing provisions—e.g., multiple damages, fee shifts—increases the ideological distance between Congress and President); Jerry L. Mashaw, Prodelegation: Why Administrators Should Make Political Decisions, 1 J.L. Econ. & Org. 81, 95–99 (1985) (offering classic account of legislative delegation to achieve context-sensitive policy implementation). 60. See, e.g., Jeb Barnes, Overruled? Legislative Overrides, Pluralism, and Contemporary Court-Congress Relations 6, 45 (2004) (sketching, as part of broader theoretical framework, rosy, “pluralistic” view of political overrides of judicial decisions as “salutary form of inter-branch feedback”). For a version of Barnes’s less cheerful, “hyper- pluralistic” view, see infra notes 77–87 and accompanying text. 61. See supra notes 12, 54 and accompanying text (elaborating on innovation in litigation). 62. See supra notes 40–43 and accompanying text (recounting changes in plaintiffs’ bar and effect on litigation frequency). 1934 COLUMBIA LAW REVIEW [Vol. 114:1913 better able to bring larger cases against better-resourced defendants.63 Yet more sophisticated and better-capitalized firms will not just hunt big- ger game. They will also possess both the intellectual wherewithal to mint new theories of liability and the resources to deploy and defend those theories in protracted trial and appellate proceedings.64 Here again, private enforcement efforts may exhibit predictable patterns over time, with substantial legal innovation emerging only after sufficient private enforcement capacity has come online. This richer temporal perspective points toward further and critically important insights about the pathways down which private enforcement efforts are likely to travel. For instance, increasingly sophisticated and well-resourced private enforcers will tend to push into statutory and regulatory interstices.65 Part of this is a straightforward extension of deterrence theory: Competent and “well-incentivized private enforcers will, to at least some extent, secure their own obsolescence as regulatory targets adapt to a heightened risk of detection and punishment.”66 Growth in the sophistication and resources of private enforcers will also tend to make them less risk averse—that is, more risk preferring—and thus more willing to initiate lower-probability (but higher-payout) cases than before.67 Combining these dynamics, one might expect that private enforcers will, as a litigation regime matures, shift from targeting rela-

63. See Coffee, Understanding, supra note 40, at 704–06 (analyzing effect of firm size and resources on risk preferences and litigation capacity); see also Howard M. Erichson, Coattail Class Action: Reflections on Microsoft, Tobacco, and the Mixing of Public and Private Lawyering in Mass Litigation, 34 U.C. Davis L. Rev. 1, 3 n.5 (2000) (suggesting increase in plaintiffs’ bar’s use of winnings in asbestos and mass-tort cases to invest in new litigation initiatives). 64. See Thomas F. Burke, On the Rights Track: The Americans with Disabilities Act, in Comparative Disadvantages? Social Regulations and the Global Economy 242, 246 (Pietro S. Nivola ed., 1997) (“[P]otential plaintiffs with resources can afford to make tenuous claims that stretch the law.”); Park, supra note 12, at 159 (“Class action attorneys are entrepreneurial enforcers, willing to invest resources in pursuing innovative theories of wrongdoing.”). 65. This Article’s use of the terms “interstices” and “interstitial” is meant to describe ambiguous legal mandates without reference to the centrality or importance of the provision. To that extent, it breaks somewhat from the recent use of the term in administrative law implying that a provision is less central to a scheme. See, e.g., Barnhart v. Walton, 535 U.S. 212, 222 (2002) (considering “interstitial nature of the legal question” as one factor in applying Chevron deference test). See generally Barnes, supra note 60, at 39 (using similar operationalization in which even interstitial questions can implicate “high-stakes policy issues”). 66. Engstrom, Harnessing, supra note 18, at 1323; see also Margaret H. Lemos & Alex Stein, Strategic Enforcement, 95 Minn. L. Rev. 9 (2009) (reviewing deterrence theory); Wendy Naysnerski & Tom Tietenberg, Private Enforcement of Federal Environmental Law, 68 Land Econ. 28, 43 (1992) (“Since the rise of private enforcement increases the likelihood that violations will be detected and prosecuted, it should increase the observed degree of compliance with the regulations . . . . Successful citizen suits ultimately undermine the very reason for their existence.”). 67. See Coffee, Understanding, supra note 40, at 705–06 (stating plaintiff’s attorney is “repeat player and therefore less likely to be risk averse”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1935 tively clear to relatively ambiguous segments of the statutory and regulatory code as a growing corps of private attorneys general makes sanction for breach of the former a virtual certainty.68 More specifically, one might expect that private litigation efforts will, over time, tend to move away from targeting relatively clearer rule-based formulations of legal mandates and toward more open-ended standard-based formulations.69 Over time, this evolution will work a subtle but fundamental change in the character of litigation efforts. As more sophisticated and better- capitalized private enforcers focus their energies on more ambiguous standard-based legal mandates, enforcement efforts will grow in scale and scope, thereby exhibiting an increasingly “regulatory” cast. In particular, one might expect that a growing cadre of well-resourced private enforcers working within legal interstices will increasingly bring actions targeting multiple defendants engaged in common industry practices that skirt legality, rather than targeting solo malefactors engaged in patently illegal conduct. When policymakers deploy litigation as a regulatory tool, private enforcers may not just be patient, dutiful sentinels moving against isolated bouts of illegality. Rather, private enforcers can wield agenda-setting and lawmaking power, deploying a version of what critics have dubbed “regulation by litigation” to set industry-wide rules via litigated judgments and settlements that might otherwise be achieved via legislation or administrative rulemaking.70

68. This claim depends on a plausible, but hardly ironclad, assumption that defendant-side risk preferences remain constant. A complicating factor is an influential line of theory suggesting that legal ambiguity is as likely to induce overcompliance as it is undercompliance with regulatory mandates relative to the social optimum. See John E. Calfee & Richard Craswell, Some Effects of Uncertainty on Compliance with Legal Standards, 70 Va. L. Rev. 965, 965 (1984) (discussing how legal uncertainty incentivizes parties to modify behavior to greater extent than rule requires); Richard Craswell & John E. Calfee, Deterrence and Uncertain Legal Standards, 2 J.L. Econ. & Org. 279, 280 (1986) (concluding overcompliance is common where uncertainty is relatively small and undercompliance is more likely where uncertainty is broad). This suggests that, for at least at some levels of misconduct—for instance, conduct that barely crosses the threshold of illegality—there may be fewer enforcement opportunities where legal mandates are relatively ambiguous. Additional work might attempt to specify the conditions under which aggregate overcompliance will sufficiently swamp undercompliance such that private enforcement does not flow toward less determinate legal mandates. 69. The most convincing proof of this dynamic is situations involving the reverse of the hypothesized shift: the striking decline in private enforcement efforts when more open-ended, standard-based regulatory mandates are, via judicial or administrative elaboration, made more rule-like. See, e.g., Mark A. Sargent, The New Regulation D: Deregulation, Federalism and the Dynamics of Regulatory Reform, 68 Wash. U. L.Q. 225, 226–27 (1990) (noting decline of private litigation efforts following promulgation of so- called Regulation D, providing safe harbor for certain nonpublic securities offerings). 70. See Andrew P. Morriss, Bruce Yandle & Andrew Dorchak, Regulation by Litigation 1 (2009) (defining “regulation-by-litigation” as “bringing suits and achieving ends that could be and traditionally had been achieved by regulatory agencies using rulemaking procedures”); Patrick Luff, Risk Regulation and Regulatory Litigation, 64 Rutgers L. Rev. 73, 89–95 (2011) (discussing competing definitions of “regulatory 1936 COLUMBIA LAW REVIEW [Vol. 114:1913

Of course, rising enforcement scale and scope need not be a bad thing. Privately driven efforts to update statutes may, as one might view the sexual-harassment example just noted, prove to be desirable, welfare- improving policies.71 Nor is a tendency toward larger-scale, wider-scope litigation efforts unique to private enforcement. Indeed, one might expect a similar trend in regimes delegating enforcement authority solely to prosecutors and agencies as they build enforcement capacity and acquire needed expertise. It should therefore be unsurprising that critics level the “regulation by litigation” critique at public enforcement efforts as well.72 But note a further implication that extends from private enforcement’s tendency to push into legal interstices: Privately driven legal innovation may, contrary to the more optimistic view of private enforcement, defy meaningful political, democratically accountable control. To be sure, the mechanisms here may not be obvious at first glance. As previously noted, courts, agencies, and even the legislature itself stand ready to check or override private litigation efforts that stray beyond statutory purposes, thus limiting any “drift.”73 Concrete examples litigation” and outlining important characteristics); William M. Sage, Unfinished Business: How Litigation Relates to Health Care Regulation, 28 J. Health Pol. Pol’y & L. 387, 389, 392–93 (2003) (discussing possible problems with use of litigation to fill regulatory gaps in healthcare context); Christopher D. Zalesky, Pharmaceutical Marketing Practices: Balancing Public Health and Law Enforcement Interests; Moving Beyond Regulation- Through-Litigation, 39 J. Health L. 235, 241 & n.27 (2006) (discussing impact of DOJ and HHS law enforcement efforts and infrequency of FCA cases pursued to verdict). For other seminal contributions, see Robert A. Kagan, Adversarial Legalism: The American Way of Law 3 (2001) (introducing concept of “adversarial legalism” as “policymaking, policy implementation, and dispute resolution by means of lawyer-dominated litigation”); Regulation Through Litigation 1 (W. Kip Viscusi ed., 2002) (discussing interaction between regulation and litigation). 71. At least one prominent theory of statutory interpretation is built around the possibility of “judicial updating” of obsolete statutes in this manner. See William N. Eskridge, Jr., Dynamic Statutory Interpretation 5–6 (1994) (advancing theory of dynamic statutory interpretation). For earlier entrants to the literature, see Guido Calabresi, A Common Law for the Age of Statutes 2 (1982) (discussing “legal obsolescence” and role of judiciary in addressing obsolete law); Henry J. Friendly, The Gap in Lawmaking—Judges Who Can’t and Legislators Who Won’t, 63 Colum. L. Rev. 787, 792 (1963) (discussing lawmaking roles of judge and legislator); Roscoe Pound, Anachronisms in Law, 3 J. Am. Judicature Soc’y 142 (1919) (discussing obsolete doctrines and statutes as impediment to justice). 72. See Morriss et al., supra note 70, at 1 (profiling public enforcement initiatives, such as EPA litigation against heavy-duty diesel-engine manufacturers, alongside private enforcement efforts, such as asbestos litigation); see also Roberta S. Karmel, Regulation by Prosecution 95 (1982) (noting SEC’s “predilection for formulating regulatory policy through the prosecution of enforcement cases”); Michael I. Krauss, Regulation Masquerading as Judgment: Chaos Masquerading as Tort Law, 71 Miss. L.J. 631, 633 (2001) (arguing against use of government suits to accomplish regulation). 73. See supra note 58 and accompanying text (emphasizing multiple veto gates in private litigation model). Admittedly, this Article’s use of the “drift” concept raises a baseline question: Is “drift” to be measured relative to the preferences of the enacting 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1937 abound, suggesting one should take care not to overstate political- or democratic-control concerns. Thus, the Supreme Court’s recent Wal- Mart Stores, Inc. v. Dukes decision dashed the hopes of the employment plaintiff bar to make excessive delegation of managerial discretion into actionable job discrimination under Title VII.74 Similarly, courts and state legislatures alike stymied the efforts of the personal-injury bar to establish “enabling torts” as a way to reach deeper-pocketed defendants.75 As a final note, recall that drift via privately driven legal innovation may be uncontroversially good and, even if not, may have factored in the legislature’s decision to deploy private litigation over regulatory alternatives in the first place.76 To that extent, perhaps one should not think in “drift” terms at all, but rather salutary “evolution” that is entirely within congressional cognizance. Still, theory and evidence suggest that drift, whether or not within legislative contemplation at enactment, may be substantial. Part of the reason is the fragmented supermajoritarian structure of the American state and the limited institutional will and capacity of courts, legislatures, and agencies, which can combine to create a large discretion space—or, in political-science argot, a “gridlock region”—within which private

Congress or to those of the current Congress (and veto-holding President)? Compare, e.g., William M. Landes & Richard A. Posner, The Independent Judiciary in an Interest-Group Perspective, 18 J.L. & Econ. 875, 885–87 (1975) (advancing theory in which courts strive to enforce initial legislative bargain struck), with Einer Elhauge, Statutory Default Rules: How to Interpret Unclear Legislation 9–10 (2008) (distinguishing “enacting legislative” preferences from “current legislative” preferences and building theory of statutory interpretation keyed to latter). Note, however, that nothing turns on this distinction for this Article’s purposes, as the mechanisms the Article develops can drive deviations from either set of preferences. 74. See 131 S. Ct. 2541, 2553–55 (2011) (rejecting “social framework” analysis that would have rendered Wal-Mart liable for excessively delegating discretion over hiring and promotion to local managers); Suzette M. Malveaux, How Goliath Won: The Future Implications of Dukes v. Wal-Mart, 106 Nw. U. L. Rev. Colloquy 34, 52 (2011), http://www. law.northwestern.edu/lawreview/colloquy/2011/18/LRColl2011n18Malveaux.pdf (on file with the Columbia Law Review) (“Dukes has redefined the class certification requirements for Title VII cases in ways that jeopardize potentially meritorious challenges to systemic employment discrimination.”). 75. See John C. P. Goldberg & Benjamin C. Zipursky, Tort Law and Responsibility, in Philosophical Foundations of the Law of Torts 17, 23–25 (John Oberdiek ed., 2014) (discussing circumscribed expansion of liability for background actor negligence in American courts); cf. Robert L. Rabin, Enabling Torts, 49 DePaul L. Rev. 435, 447 (1999) (noting judicial rejection of “enabling tort” theory of recovery). Another example is the Supreme Court’s rejection of the plaintiffs’ securities bar’s effort to establish enterprise and gatekeeper liability. See Stoneridge Inv. Partners, LLC v. Scientific-Atl., Inc., 552 U.S. 148, 153 (2008) (holding securities laws generally do not reach defendants’ vendors or customers); Cent. Bank of Denver v. First Interstate Bank of Denver, 511 U.S. 164, 191 (1994) (holding securities laws provide no private right of action against aiders and abettors). 76. See supra note 59 and accompanying text (highlighting logic of legislative delegation of enforcement authority to private litigants). 1938 COLUMBIA LAW REVIEW [Vol. 114:1913 litigants are only imperfectly monitored and checked.77 More concretely, judges, in addition to having policy preferences of their own, possess neither policy-specific expertise nor a synoptic view of the enforcement landscape, sharply limiting their ability to gauge how a given legal innovation urged upon them maps onto legislative policy aims.78 The tendency of repeat players within litigation regimes to “play for rules” by settling bad cases and litigating good ones further muddies judges’ views.79 Legislators, too, may prove unreliable overseers, even of their own legislative bargains. This is not just because they cannot act on their preferences—though, as just noted, the American separation-of-powers system, with its multiple veto gates, makes legislative inertia a distinct possibility.80 Rational legislators may also simply lack any preferences apart from staying above the fray.81

77. See generally Keith Krehbiel, Pivotal Politics: A Theory of U.S. Lawmaking (1998) (offering classic account of “gridlock regions” within separation-of-powers systems). This, of course, is the other side of the coin of models, noted previously, in which separated powers coordinate to keep policy within certain bounds. See supra note 58 and accompanying text (noting classic Eskridge and Ferejohn model). 78. See Frederick Schauer & Richard Zeckhauser, The Trouble with Cases, in Regulation Versus Litigation: Perspectives from Economics and Law 45, 46–59 (Daniel P. Kessler ed., 2011) (“[T]he goals and incentives of the litigation process are likely to contribute to aberrational rather than representative cases being the subject of lawsuits . . . caus[ing] the policy that emerges from litigation to be systematically based on an imperfect picture of the terrain that the policy is designed to regulate.”); see also Gillian K. Hadfield, Bias in the Evolution of Legal Rules, 80 Geo. L.J. 583, 594–96 (1992) (noting skeptical view that judges will receive systematically skewed information about regulatory realities from litigated cases). 79. See Marc Galanter, Why the “Haves” Come Out Ahead: Speculations on the Limits of Legal Change, 9 Law & Soc’y Rev. 95, 97–103 (1974) [hereinafter Galanter, Why the “Haves” Come Out Ahead] (discussing repeat-play dynamics and implications for rule development). In a purely public enforcement regime, government is the ultimate repeat player. Inclusion of a private enforcement mechanism, however, adds in one-shotters whose lack of a strategic perspective beyond a given case allows repeat-play defendants in particular to systematically alter legal development. 80. See Matthew R. Christiansen & William N. Eskridge, Jr., Congressional Overrides of Supreme Court Statutory Interpretation Decisions, 1967–2011, 92 Tex. L. Rev. 1317, 1320, 1427 (2014) (examining congressional overrides and finding “restorative” overrides repudiating judicial interpretation as misrepresenting legislative intent are far outnumbered by overrides “updating” or “clarifying” policy); William N. Eskridge, Jr., Dynamic Statutory Interpretation, 135 U. Pa. L. Rev. 1479, 1525 (1987) (arguing Congress will “only occasionally and adventitiously . . . respond to judicial statutory interpretations at odds with original intent or purpose”); William N. Eskridge, Jr., Overriding Supreme Court Statutory Interpretation Decisions, 101 Yale L.J. 331, 335–36 (1991) (finding Congress overrode Supreme Court statutory decisions roughly a dozen times per year between 1975 and 1990). See generally Barnes, supra note 60, at 42–43 & nn.56–58 (cataloguing scholarship on legislative “overrides” of judicial decisions and noting “daunting political obstacles” override efforts face, including legislative “inertia”). 81. See Barnes, supra note 60, at 67–68 (sketching “hyperpluralist” theory in which Congress does not reliably pass clarifying overrides of judicial interpretations and instead enacts vague measures, allowing it to engage in credit-claiming and blame-avoidance behavior). See generally David Mayhew, Congress: The Electoral Connection (2d ed. 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1939

In comparison, one might consider agencies better positioned than courts or legislatures to police privately driven legal innovations and ensure their fidelity to the legislative design. As previously noted, agencies with rulemaking powers can override deviant statutory interpretations by promulgating contrary regulations.82 Agencies can also be vested with formal “gatekeeper” powers authorizing them to take over control of private lawsuits or dismiss them outright.83 Yet agencies armed with either power may, like judges and legislators, be imperfect guardians of statutory elaboration. One reason is the usual concern, already noted, about regulatory “capture.”84 A subtler concern holds that politically conscious agencies will allocate resources with an eye to collecting political rewards and ensuring the continued flow of resources to the agency, potentially compromising the agency’s ability to play an innovation-policing role.85 In particular, risk-averse agency officials may be disinclined to block novel private enforcement efforts using their rulemaking or gatekeeper powers where subsequent events could turn up evidence of widespread wrongdoing, thus embarrassing the agency.86 This tendency to underpolice private enforcement efforts is exacerbated by the presence of a judicial backstop, yielding a kind of “bailout effect”

2004) (offering seminal analysis of legislative position-taking and blame-avoidance incentives). 82. See supra note 58 and accompanying text (discussing institutional veto gates in separation-of-powers model). Importantly, an agency can promulgate rules that reverse an innovation even where a court has already declared the innovation to be a permissible statutory construction. See Nat’l Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 982 (2005) (holding agency statutory interpretation is entitled to deference in face of prior judicial interpretation so long as court did not determine its own construction to be sole permissible interpretation). And even where an agency lacks rulemaking authority, it can participate as an amicus curiae, urging adoption of the government’s interpretation. See Arthur F. Greenbaum, Government Participation in Private Litigation, 21 Ariz. St. L.J. 853, 872–76 (1989) (discussing government agency participation in private litigation as amicus curiae). 83. See Engstrom, Gatekeepers, supra note 16, at 619–20 (discussing “gatekeeping” powers of agencies). 84. See supra note 31 and accompanying text (discussing capture concept). 85. See David Freeman Engstrom, Public Regulation of Private Enforcement: Empirical Analysis of DOJ Oversight of Qui Tam Litigation Under the False Claims Act, 107 Nw. U. L. Rev. 1689, 1700–06 (2013) [hereinafter Engstrom, Public Regulation] (detailing this concern in gatekeeper context). On the general tendency of agencies to pursue certain observable bureaucratic outputs in order to satisfy political overseers, see James Q. Wilson, Bureaucracy: What Government Agencies Do and Why They Do It 75–83 (1991). 86. This is consistent with the common view that agencies will be excessively cautious, defensive, and “scandal-minimizing” within their regulatory bailiwicks. See Christopher C. DeMuth & Douglas H. Ginsburg, White House Review of Agency Rulemaking, 99 Harv. L. Rev. 1075, 1080 (1986) (noting “excessively cautious” nature of regulators); James Q. Wilson, The Politics of Regulation, in The Politics of Regulation 378 (James Q. Wilson ed., 1980) (describing government agencies as having “defensive, threat-avoiding, scandal- minimizing instincts”). 1940 COLUMBIA LAW REVIEW [Vol. 114:1913 in which agencies rationally shift the cost of regulating private lawsuits, whether actual or reputational, to the courts.87 Yet even if the “discretion” or “gridlock” space within which private litigation operates is small, there is a second, and potentially more important, reason to credit drift concerns: Privately driven legal innovation will often be incremental in ways that can frustrate and even defeat political- control efforts. Some of the logic here harkens back to the limited will and capacity of legislatures, courts, and agencies. For instance, judicial fallibility and the lack of interjurisdictional stare decisis will yield interpretive patchworks that allow litigation asserting a novel liability theory to go forward in one court even after its invalidation in another.88 Moreover, even where legislators and agency officials are willing to act, their efforts will often be sluggish. Legislators are busy, and the legislative process is slowed by multiple veto points.89 Agencies are subject to oner- ous administrative procedures that, though reducing the democratic deficit when unelected bureaucrats make policy, can also “ossify” agency action and slow rulemaking to a crawl.90 As a result, political override of privately driven legal innovations will rarely be rapid, let alone instanta- neous.91 Worse, the need for override may not be apparent until litigation outcomes are irreversible, once final judgments have been issued or binding settlements have been inked.92 The sluggish response of legislators and agencies to litigation efforts is particularly important, for privately driven legal innovations are not just the one-off product of litigation struggles or mere collateral damage within the system. Litigation efforts also reshape the identities, interests,

87. A similar dynamic has been noted in the context of judicial review. See, e.g., Mark Tushnet, Taking the Constitution Away from the Courts 57–58 (1999) (arguing “judicial overhang” can distort legislative behavior); Justin Fox & Matthew C. Stephenson, Judicial Review as a Response to Political Posturing, 105 Am. Pol. Sci. Rev. 397, 397 (2011) (describing “bailout effect” in which “judicial review may rescue elected officials from the consequences of ill-advised policies”). 88. See supra note 78 and accompanying text (discussing institutional factors limiting judges’ ability to evaluate policy effects of novel litigation theories). Furthermore, the Supreme Court’s limited docket will often allow splits to persist. See Peter L. Strauss, One Hundred Fifty Cases Per Year: Some Implications of the Supreme Court’s Limited Resources for Judicial Review of Agency Action, 87 Colum. L. Rev. 1093, 1100–05 (1987) (noting Court’s “failure to choose a case may have large implications for the coherence and uniformity of the body of law for which the Court is responsible”). 89. See supra note 80 and accompanying text (discussing problem of legislative inertia). 90. See Richard J. Pierce, Jr., Seven Ways to Deossify Agency Rulemaking, 47 Admin. L. Rev. 59, 60–62 (1995) (discussing “ossification” problem). 91. See, e.g., Donald C. Langevoort & Robert B. Thompson, “Publicness” in Contemporary Securities Regulation After the JOBS Act, 101 Geo. L.J. 337, 342 (2013) (noting tendency of private enforcers to fill unregulated space before SEC can meaningfully respond via rulemaking). 92. See, e.g., Plaut v. Spendthrift Farm, Inc., 514 U.S. 211, 227 (1995) (“Congress may not declare by retroactive legislation that the law applicable to that very case was something other than what the courts said it was.”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1941 and capacities of potential political actors.93 Large paydays provide the plaintiffs’ bar with a war chest with which to protect its hard-fought litigation gains through the legislative and administrative process.94 And on the defense side, a judgment against a large industry actor may insulate it from further legal attack, whether through the operation of preclusion doctrine or because the defendant alters its organizational routines in difficult-to-reverse ways to avoid further legal entanglement.95 This can dampen such an actor’s incentives to join industry lobbying efforts to reverse a given legal innovation and, indeed, may create incentives to disrupt such efforts to narrow the competitive advantages enjoyed by industry members who have not yet faced litigation.96 To return once more to the job-discrimination context, had Wal-Mart suffered a large judgment in the Dukes case, the company would have faced powerful incentives to ensure that its big-box competitors were likewise subject to large-scale class actions asserting discrimination via excessive managerial discretion.97 The result is a potentially powerful dynamic in which private litigants do not just enjoy relatively free rein within a discretion space. Rather,

93. See Barnes, supra note 60, at 42 (“[S]tatutory interpretation decisions may break apart legislative coalitions.”). Further versions of this argument draw from a growing political-science literature on how policy enactments can reshape the interest-group environment, fragmenting “issue networks” in ways that place cross-pressures on elected officials and make policy repeal, or even amendment, unlikely. See generally Living Legislation: Durability, Change, and the Politics of American Lawmaking (Jeffery A. Jenkins & Eric M. Patashnik eds., 2012) (compiling work of authors further developing this view). 94. Cf. Barnes, supra note 60, at 40 (noting entrepreneurial lawyers “have proven both resourceful litigants, who find new ways to use lawsuits to pursue their political agendas, and powerful lobbyists, who reinforce and expand the role of courts and litigation in American policy-making”). 95. A long literature documents how employers altered bureaucratic routines in response to job-discrimination suits. See, e.g., Frank Dobbin & John R. Sutton, The Strength of a Weak State: The Rights Revolution and the Rise of Human Resources Management Divisions, 104 Am. J. Soc. 441, 467 (1998) (describing increased “diffusion of personnel, benefits, health and safety, and EEO/AA offices”). But see Lauren B. Edelman, Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law, 97 Am. J. Soc. 1531, 1554 (1992) (arguing employers engaged in symbolic compliance without altering organizational practices). Many of these changes are sunk costs or otherwise entail shifts in organizational culture and norms, making reforms difficult to unwind. 96. An obvious analogy here is the suggestion by federalism scholars that states that already engage in regulation will invite federal regulation to narrow competitive advantages with states that do not. See Lynn A. Baker, Conditional Federal Spending After Lopez, 95 Colum. L. Rev. 1911, 1948–49 (1995) (noting possibility strategic states will invite conditional spending programs “to harness the federal lawmaking power to restrict the competition for residents and tax dollars that would otherwise exist among them”). 97. Cf. supra note 74 and accompanying text (citing Dukes as example of limitations on statutory drift). In theory, even a judgment favorable to Wal-Mart could create such a dynamic by incentivizing Wal-Mart to use preclusion as a shield while lobbying in favor of legal liability against competitors. 1942 COLUMBIA LAW REVIEW [Vol. 114:1913 they can redefine, and reconfigure, the bounds of the space itself. By remaking the political landscape, private enforcement efforts may, in the jargon of political science and economics, yield feedback effects, path dependencies, and an increasing-returns process wherein early litigation outcomes render privately driven legal innovations surprisingly robust even where, viewed ex ante, they plainly conflict with legislative preferences.98 Over time, private enforcement may thus drive legal mandates in different directions than one might expect if enforcement authority were in purely public hands.

** *

The above analysis raises a raft of issues, which cannot reasonably be addressed in a single study. Rather, the goal has been to distill disparate scholarly literatures into a manageable set of questions that can be brought to the data. An initial pair of questions relates to litigation quantity: Are litigation flows responsive to shifts in litigation incentives, particularly legislated ones, suggesting litigants can efficiently assimilate information about their likelihood of success? Or do litigation flows instead exhibit “gold rush” tendencies or lurching behavior more generally, suggesting inefficient equilibration in response to shifting incentives? A second pair of questions concerns litigation’s character: Is the character of private enforcement efforts relatively static over time? Or do private enforcement efforts tend, as theory suggests, to grow in scale and regulatory scope over time, pushing into regulatory interstices and exploiting regulatory ambiguities in suits targeting much or all of an industry? A third and final pair of questions concerns political control: Are private litigation efforts that deviate from legislative purposes subject to meaningful control by courts, agencies, or the legislature itself? Or will an increasingly sophisticated corps of private enforcers pressing new liability theories drive the elaboration of legal mandates in directions they would not travel in the absence of a private enforcement mechanism? Some of these questions—those regarding litigation flows, or trends in the scale and scope of litigation over time—are conducive to quantitative testing. Other questions—those involving public control and whether private enforcement uniquely drives legal evolution—raise complex counterfactuals likely to produce only suggestive answers, and only after

98. See Paul Pierson, Increasing Returns, Path Dependence, and the Study of Politics, 94 Am. Pol. Sci. Rev. 251, 251–53 (2000) (offering temporal theory of politics in which highly contingent events occurring early in process feed back into future choices, making multiple different outcomes possible). See generally Paul Pierson, Politics in Time: History, Institutions, and Social Analysis (2004) (further developing these views). For an illuminating general analysis of different versions of the path-dependency concept and their application to legal development, see generally Oona A. Hathaway, Path Dependence in the Law: The Course and Pattern of Legal Change in a Common Law System, 86 Iowa L. Rev. 601 (2001). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1943 sustained qualitative engagement with the actual workings of a regime. Yet few scholarly efforts attempt either type of analysis, at least in part because of the difficulty of finding a self-contained and fully observable litigation context in which to test key hypotheses.99 The next three Parts begin to remedy the resulting gap in scholarly understanding by training an empirical eye on a relatively new, increasingly controversial, and highly tractable type of litigation that is fueling intense debate about the quantity, character, and public control of private enforcement efforts: qui tam lawsuits brought under the FCA.

II. THE CASE OF QUI TAM LITIGATION

The False Claims Act offers a natural laboratory to test competing accounts of private enforcement’s pathways. This Part offers a brief overview of the FCA’s qui tam regime, makes the case for using the regime as a case study, and describes the data used in the empirical analysis to come in Parts III and IV.

A. Qui Tam Basics The modern version of the False Claims Act dates to 1986 when Congress, moved by growing public concern about defense-procurement fraud, rescued the Civil War–era statute from decades of near desuetude by strengthening several of its core provisions.100 In its current form, the FCA creates liability for any person who knowingly submits a false money claim to the government, uses a false statement to induce the government to pay a false claim, conspires to defraud the government into paying a false claim, or uses a false statement to reduce an obligation to pay money to the government.101 Penalties are steep, with fines ranging from $5,500 to $11,000 per false claim, plus treble the amount of any

99. As noted previously, a growing empirical literature tests litigant responsiveness to shifts in legislated incentives. See supra note 11 and accompanying text. But because of confidential settlements such research is limited by the lack of information about case outcomes. See Laura Beth Nielsen et al., Individual Justice or Collective Legal Mobilization? Employment Discrimination Litigation in the Post Civil Rights United States, 7 J. Empirical Legal Stud. 175, 187 (2010) (noting settlement outcomes “typically are unavailable because of confidentiality agreements”); see also Catherine T. Harris et al., Does Being a Repeat Player Make a Difference? The Impact of Attorney Experience and Case-Picking on the Outcome of Medical Malpractice Lawsuits, 8 Yale J. Health Pol’y L. & Ethics 253, 263 (2008) (“[I]n practice, many settlement amounts are kept confidential.”); Minna J. Kotkin, Outing Outcomes: An Empirical Study of Confidential Employment Discrimination Settlements, 64 Wash. & Lee L. Rev. 111, 118 (2007) (observing “[s]uccessful claims are made invisible by secret settlements” and attempting study of settlement outcomes). 100. See United States ex rel. Springfield Terminal Ry. Co. v. Quinn, 14 F.3d 645, 649–51 (D.C. Cir. 1994) (recounting substance of 1986 amendments). 101. See 31 U.S.C. § 3729(a) (2012). 1944 COLUMBIA LAW REVIEW [Vol. 114:1913 proven damages.102 While the DOJ can initiate either criminal or civil actions against fraudsters, in practice most FCA enforcement efforts are initiated as private lawsuits brought pursuant to the FCA’s qui tam provisions.103 Those provisions authorize private persons, dubbed “relators,” to sue private parties by alleging fraud against the United States. These relators earn a bounty equal to a portion—ranging from fifteen to thirty percent—of any (trebled) recovery.104

FIGURE 2: QUI TAM FILINGS BY CASE TYPE AND TOTAL RECOVERIES, 1986–2013

Figure 2 shows rapid growth in qui tam litigation since 1986 such that the regime, as noted previously, now rivals or even eclipses securities and antitrust in annual filings and recoveries.105 The 1986 amendments have also spurred a rich diversity of claim types. Continuing with Figure 2, the most common of these assert fraud in connection with federally funded healthcare services under Medicare, Medicaid, and defense- procurement contracts. But roughly one-quarter take “Other” forms and cover the waterfront of federal programs and expenditures. Qui tam claims now regularly target underpayment of oil and gas royalties owed

102. Id.; see also The False Claims Act: A Primer, U.S. Dep’t of Justice, http://www. justice.gov/civil/docs_forms/C-FRAUDS_FCA_Primer.pdf (on file with the Columbia Law Review) (last visited Sept. 29, 2014) (explaining current adjusted penalties range from $5,500 to $11,000). 103. See 31 U.S.C. § 3730(a)–(b) (outlining authority of DOJ and private persons to bring suit); see also Fraud Statistics, supra note 6 (reporting more than 700 qui tam suits but only ninety-three government-initiated FCA “matters” in 2013). 104. 31 U.S.C. § 3730(b)–(d). 105. See supra note 6 and accompanying text (illustrating rapid growth of qui tam litigation). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1945 when defendants extract natural resources from federal lands. They also reach a myriad of other frauds in connection with federally insured education and housing loans, federal research grants, and federal funds distributed in connection with Hurricane Katrina relief efforts and the Troubled Asset Relief Program.106

B. The Case for Qui Tam as a Case Study While qui tam’s dramatic rise makes study of its basic workings important in its own right, the regime also offers an ideal vehicle for empirical exploration of some of Part I’s theoretical predictions about private enforcement’s pathways. First, the FCA regime epitomizes legislative use of “private attorneys general” to achieve regulatory ends and the institutional-design challenges legislators face in doing so. For instance, an initial set of FCA provisions shapes the pool of relators who are eligible to recover qui tam bounties and thus aims to minimize “parasitic” qui tam suits that do not offer the government information it does not already have or merely piggyback on existing public enforcement efforts.107 A further set of provisions aims to shape private enforcement efforts by granting the Attorney General—and, by further delegation, the DOJ’s Civil Fraud Division—substantial authority to oversee and control qui tam lawsuits. By far the most significant mechanism of public control is the DOJ’s power to intervene in qui tam actions, taking primary control over their prosecution.108 The DOJ enjoys other control rights as well. Thus, even where DOJ declines to intervene, it may still dismiss or settle a qui tam case out from under a private relator entirely, subject only to a

106. For a comprehensive overview of claim types, see Common Types of Qui Tam Cases, Phillips & Cohen LLP, http://www.phillipsandcohen.com/Common-Types-of-Qui- Tam-Cases/ (on file with the Columbia Law Review) (last visited Sept. 15, 2014). 107. A pair of provisions jurisdictionally bars claims mirroring a previously filed suit or an existing government-enforcement proceeding. 31 U.S.C. § 3730(b)(5), (e)(3). A further provision bars claims that have been “publicly disclosed” in the press or a government report except where the relator is an “original source”—defined as having voluntarily disclosed the information before public disclosure or having independent knowledge—of the information on which the allegation is based. 31 U.S.C. § 3730(e)(4)(A)–(B); see also Schindler Elevator Corp. v. United States ex rel. Kirk, 131 S. Ct. 1885, 1894 (2011) (construing “public disclosure” provision); Rockwell Int’l Corp. v. United States, 549 U.S. 457, 476 (2007) (construing “original source” provision). 108. More concretely, the FCA directs qui tam relators to file complaints under seal, serving copies on the DOJ. A sixty-day seal period follows during which DOJ lawyers investigate the allegations and decide whether to (i) terminate or settle the case; (ii) intervene and take “primary responsibility” over the litigation, including the ability to limit a relator’s procedural rights; or (iii) decline to intervene and allow the relator to proceed alone. See 31 U.S.C. § 3730(b)(2), (c)(1), (c)(2)(A)–(C) (providing government’s procedural options for responding to qui tam actions). Importantly, the bounty paid to a successful relator turns, at least in part, on the DOJ’s case-election decision, with the relator keeping fifteen percent to twenty percent where the DOJ joins and twenty-five to thirty percent where the DOJ declines. 31 U.S.C. § 3730(d)(1)–(2). 1946 COLUMBIA LAW REVIEW [Vol. 114:1913 basic-fairness hearing.109 Similarly, the DOJ must consent to any private dismissal or settlement of a qui tam action.110 A second way qui tam presents an ideal vehicle for exploring private enforcement’s pathways is its unique empirical tractability. Because the FCA’s 1986 revival came in the modern litigation era, case docket information, much of it electronic, is available across the lifespan of the regime. Even better, the DOJ’s statutory oversight duties mean that qui tam case outcomes, including litigated judgments and settlements (the latter of which often remain confidential in other litigation areas), are cen- trally tracked by the DOJ and discoverable via Freedom of Information Act (FOIA) requests. The result is a level of empirical observability that is unachievable in other often studied litigation contexts such as torts or job discrimination.111 Third and finally, qui tam’s rapid growth has occasioned heated debate about the volume, nature, and public control of qui tam enforcement efforts, providing a rich—if mostly anecdotal—background against which to test Part I’s theoretical predictions. One flashpoint of debate is whether there is too much or too little qui tam litigation. Defenders of the regime point out that qui tam enforcement activity is dwarfed by total federal spending or reasonable estimates of the amount of fraud in specific programmatic areas,112 implying that the problem is not too much qui tam enforcement but rather too little.113 Others, however, use

109. The Ninth Circuit, for instance, requires that the DOJ show a “rational relation” between dismissal and a valid government purpose. See United States ex rel. Sequoia Orange Co. v. Baird-Neece Packing Corp., 151 F.3d 1139, 1145 (9th Cir. 1998). 110. See 31 U.S.C. § 3730(c)(2)(A)–(B) (permitting settlement by government by involving court in both dismissal and settlement proceedings). From a procedural perspective, this power is critical, since any judgment entered for or against the relator will—because the relator stands in the government’s shoes—also preclude the government’s later assertion of transactionally related claims. This creates powerful incentives for defendants to trade a wide release of liability for a larger settlement pot. See, e.g., Searcy v. Philips Elecs. N. Am. Corp., 117 F.3d 154, 160 (5th Cir. 1997) (noting “danger that a relator can boost the value of settlement by bargaining away claims on behalf of the United States”). 111. See Harris et al., supra note 99, at 262–63 (noting difficulty “in determining case outcomes with precision” in studying tort system); Laura Beth Nielsen et al., Individual Justice or Collective Legal Mobilization? Employment Discrimination Litigation in the Post Civil Rights United States, 7 J. Empirical Legal Stud. 175, 187 (2010) (noting data on settlement size are often unavailable because of confidentiality agreements, making it difficult to draw inferences about litigation regime’s workings). 112. See Financial Crimes Report to the Public 2007, FBI, http://www.fbi.gov/stats- services/publications/fcs_report2007 (on file with the Columbia Law Review) (last visited Sept. 15, 2014) (reporting between three and ten percent of $2 trillion in healthcare spending is lost to fraud). 113. See, e.g., John L. Sinatra, Jr., Billions of Dollars Lost to Fraud Every Year, Much of It Unreported, Whistleblower Blog (Apr. 29, 2013), http://blog.thewhistleblower attorney.com/2013/04/29/billions-of-dollars-lost-to-fraud-every-year-much-of-it-unreport ed/ (on file with the Columbia Law Review) (noting that, even with $5 billion in recoveries in 2013, “there is still a lot of fraud left every year for whistleblowers to uncover and 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1947 those same numbers to argue that the FCA regime in its current form is ineffective at preventing fraud.114 A further avalanche of counter- commentary in Supreme Court filings, congressional testimony, and the popular and academic press cite the rising number of qui tam filings and a spate of recent outsized qui tam settlements as evidence that the regime is in the midst of an “out-of-control” and grossly inefficient litigation “explosion.”115 Much of this “explosion” in filings, many of these same commentators assert, are meritless lawsuits, suggesting a regime that is suffering from overentry and a “gold rush” of wasteful opportunism.116 A second flashpoint concerns the possibly changing nature of qui tam litigation. Paralleling Part I’s theoretical synthesis, qui tam’s defenders tout an increasingly specialized qui tam relator bar’s deepening capacity to target larger and more sophisticated frauds, developing innovative claims to reach misconduct that has escaped public regulators’ notice.117 Critics, by contrast, accuse the organized relators’ bar of report”); David Voreacos, Health-Care False Claims Cases Reap $18.3 Billion, Report, Bloomberg (Oct. 21, 2013, 7:15 PM), http://www.bloomberg.com/news/2013-10-21/ health-care-false-claims-cases-reap-18-3-billion-report.html (on file with the Columbia Law Review) (noting FCA recoveries “are a small fraction of the $2.8 trillion the U.S. spends annually on health care”); cf. Grassley Statement, supra note 7, at 10 (“[T]he False Claims Act is working, and it’s working fantastically.”). 114. This is the principal thrust of a recent report authored by, among others, former Deputy Attorney General David Ogden on behalf of the Chamber of Commerce. See Peter Hutt II et al., Fixing the False Claims Act: The Case for Compliance-Focused Reforms 1 (2013), available at http://www.instituteforlegalreform.com/uploads/sites/1/ Fixing_The_FCA_Pages_Web.pdf (on file with the Columbia Law Review) (arguing for reforms to FCA due to, among other things, unsatisfactory recovery amounts); see also Joan H. Krause, Following the Money in Health Care Fraud: Reflections on a Modern-Day Yellow Brick Road, 36 Am. J.L. & Med. 343, 355 (2010) [hereinafter Krause, Following] (making similar argument). For criticism of the Chamber report, see Grassley Statement, supra note 7, at 10–11 (criticizing report for vagueness and inconsistency). 115. See supra note 8 and accompanying text (collecting studies asserting anecdotal claims about qui tam “explosion”). 116. See supra note 8 and accompanying text (collecting studies making this claim); see also Christopher M. Alexion, Open the Door, Not the Floodgates: Controlling Qui Tam Litigation Under the False Claims Act, 69 Wash. & Lee L. Rev. 365, 404 (2012) (suggesting large proportion of qui tam suits is “likely to be frivolous” based on aggregated DOJ statistics); William E. Kovacic, Whistleblower Bounty Lawsuits as Monitoring Devices in Government Contracting, 29 Loy. L.A. L. Rev. 1799, 1855 (1996) [hereinafter Kovacic, Bounty Lawsuits] (“[T]he qui tam mechanism provides inadequate disincentives for relators to file meritless suits.”); Joan H. Krause, “Promises to Keep”: Health Care Providers and the Civil False Claims Act, 23 Cardozo L. Rev. 1363, 1383 (2002) [hereinafter Krause, Promises] (noting “meritless” lawsuits as possible cause of increases in qui tam litigation); Christina Orsini Broderick, Note, Qui Tam Provisions and the Public Interest: An Empirical Analysis, 107 Colum. L. Rev. 949, 971 (2007) (concluding from DOJ’s twenty-percent intervention rate that “number of frivolous suits is high”). 117. See, e.g., Bucy, supra note 7, at 8, 53, 58 (touting ability of private enforcement to attract “entrepreneurial legal talent”); Rich, supra note 8, at 1248 (“[H]ealth care qui tam actions have become more complex and target more sophisticated fraud.” (citing 1948 COLUMBIA LAW REVIEW [Vol. 114:1913 developing ever more creative forms of abusive litigation, targeting much or all of an industry in suits that exploit regulatory gray areas rather than surfacing plain-vanilla, “raw” frauds.118 These pan-industry suits, critics contend, run directly contrary to the legislative history of the 1986 amendments, which warned against use of the revived qui tam mechanism to bring “class actions or multiple separate suits based on identical facts and circumstances.”119 Particularly controversial are so-called “certification” claims, in which a relator alleges that a federal-funding recipient’s express or implied certification of compliance with a separate statutory or regulatory command as a condition of receiving federal funds renders “false” any subsequent claim made for those funds.120 Such

Robert D. McCallum, Jr., Assistant Att’y Gen., Remarks to the American Health Lawyers Association Meeting (Sept. 30, 2002))). 118. See supra note 9 (collecting studies arguing this claim); see also Kovacic, Bounty Lawsuits, supra note 116, at 1805–06 (“The growth of the plaintiffs’ qui tam bar and the success of some relators in obtaining substantial recoveries have provided the means and incentive to explore the application of far-reaching theories of fraud to a wider range of contractor conduct.”); Joan H. Krause, Healthcare Fraud and Quality of Care: A Patient- Centered Approach, 37 J. Health L. 161, 164 (noting increasing tendency of qui tam enforcement efforts to target “regulatory noncompliance, rather than billing misrepresentations”); David S. Torborg, The Dark Side of the Boom: The Peculiar Dilemma of Government Spoliation in Modern False Claims Act Litigation, 26 J.L. & Health 181, 182 (2013) (“Much of that explosion stems from aggressive and creative legal theories that challenge controversial industry practices or even well-known loopholes or waste in government policy.”); Robert Salcido, The Government Declares War on Qui Tam Plaintiffs Who Lack Inside Information: The Government’s New Policy to Dismiss These Parties in False Claims Act Litigation, Health Law., Oct. 2000, at 1, 3 [hereinafter Salcido, Government Declares War] (noting rising incidence of “industry-wide, multidefendant qui tam actions”). 119. See S. Rep No. 99-345, at 25 (1986), reprinted in 1986 U.S.C.C.A.N. 5266, 5290 (“[P]rivate enforcement under the civil False Claims Act is not meant to produce class actions or multiple separate suits based on identical facts and circumstances.”). 120. See Michael Holt & Gregory Klass, Implied Certification Under the False Claims Act, 41 Pub. Cont. L.J. 1, 1–2 (2011) (describing certification claims); see also Krause, Promises, supra note 116, at 1383 (“Recently, prosecutors and qui tam relators have sought to style various forms of regulatory noncompliance as actionable fraud.”); Christopher L. Martin, Jr., Reining in Lincoln’s Law: A Call to Limit the Implied Certification Theory of Liability Under the False Claims Act, 101 Calif. L. Rev. 227, 232–33 (2013) (advocating highly limited version of implied-certification theory); Dayna Bowen Matthew, A New Strategy to Combat Racial Inequality in American Health Care Delivery, 9 DePaul J. Health Care L. 793, 825 (2005) (arguing “plaintiffs have been active and creative in developing theories of recovery under the FCA” and citing false-certification claims as example). Note, however, that the use of the “certification” label to describe cases basing FCA liability on a separate statutory or regulatory provision is not universally accepted. Relator’s counsel have long asserted that the “certification” label is a meaningless and unhelpful defense-counsel creation designed to cabin liability despite the expressed intent of Congress in 1986 to include liability for “goods or services . . . provided in violation of . . . statute, or regulation.” S. Rep. No. 99-345, at 9. Several courts have agreed, rejecting the “certification” label entirely. See, e.g., United States ex rel. Hutcheson v. Blackstone, 647 F.3d 377, 385 (1st Cir. 2011) (examining origins of “certification” label and rejecting it as more likely to “obscure than clarify”); United States ex rel. Hendow v. Univ. of Phoenix, 461 F.3d 1166, 1172 (9th Cir. 2006) (“So long as the statement in question is 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1949 claims, as one colorful account in a Supreme Court amicus brief puts it, are the “theories du jour of the relator’s bar” and part of a “relentless drive” by qui tam lawyers “to stretch the [FCA’s] qui tam provisions into a private, and punitive, enforcement mechanism,” even where infractions are minor or technical and even where the separate statute does not provide for a private right of action.121 On this view, qui tam enforcement efforts threaten to colonize other entirely separate regulatory areas in direct contravention of congressional intent and purposes. A final, and related, flashpoint concerns the degree of public control over qui tam litigation, whether by Congress or by the DOJ in the exercise of its statutory oversight powers. Some cast qui tam enforcement as a stellar example of a public–private hybrid that leverages the unique capacities of each enforcement type or tout qui tam enforcers’ ability to serve a salutary anticapture or agency-forcing role where the DOJ or the defrauded agency is compromised in its ability to vindicate government interests.122 Critics, by contrast, fault Congress for not stepping in to override certain claim types or castigate the DOJ for not maintaining closer control over qui tam actions by utilizing its authority to terminate cases.123 On this latter view, the qui tam regime has become a “privately- driven system of penalizing a range of non-fraudulent conduct” that too often operates outside public control.124

knowingly false when made, it matters not whether it is a certification, assertion, statement, or secret handshake; False Claims liability can attach.”). 121. Brief Amicus Curiae of the Chamber of Commerce of the United States of America in Support of Petitioner, supra note 8, at 17, 19; see also supra note 9 (collecting criticisms about tethering FCA to other statutory and regulatory mandates). 122. See, e.g., The False Claims Act Correction Act (S. 2041): Strengthening the Government’s Most Effective Tool Against Fraud for the 21st Century: Hearing on S. 2041 Before the S. Comm. on the Judiciary, 110th Cong. 27 (2008) (testimony of John E. Clark, Of Counsel, Goode, Casseb, Jones, Riklin, Choate & Watson, P.C.) (touting ability of specialized qui tam counsel to align private enforcement efforts and public enforcement priorities through close ties with DOJ); Bucy, supra note 7, at 58 (extolling virtues of “hybrid” public–private regime and noting “formidable legal and investigative resources the FCA brings to the public regulatory efforts”). 123. See Engstrom, Public Regulation, supra note 85, at 1693 & n.16 (collecting and discussing criticisms). 124. Dayna Bowen Matthew, The Moral Hazard Problem with Privatization of Public Enforcement: The Case of Pharmaceutical Fraud, 40 U. Mich. J.L. Reform 281, 330 (2007) [hereinafter Matthew, Moral Hazard]; see also Krause, Twenty-Five, supra note 9, at 16 (noting qui tam’s critics often blame “relators, who are free to pursue virtually any theory of FCA liability—in large part because, by definition, they exist outside the governmental enforcement apparatus”); Dayna Bowen Matthew, Tainted Prosecution of Tainted Claims: The Law, Economics, and Ethics of Fighting Medical Fraud Under the Civil False Claims Act, 76 Ind. L.J. 525, 566 (2001) [hereinafter Matthew, Tainted] (“[Q]ui tam tainted claims pursued under the FCA are a significant departure from public enforcement of the anti-kickback and self-referral laws that Congress originally designed.”). 1950 COLUMBIA LAW REVIEW [Vol. 114:1913

C. Data Overview What unites the above welter of competing claims is a near-total absence of systematic empirical inquiry that tests core premises.125 To remedy this lack of empirical study, data were assembled from two main sources. First, information on the roughly 6,000 unsealed qui tam lawsuits filed since 1986 was obtained via FOIA requests served on the DOJ. The provided information includes, among other things, the judicial district in which the case was filed, filing dates, DOJ case-election dates and decisions, and case outcomes (including the amount of any recovery and the relator’s share).126 The DOJ also provided a list of sealed qui tam

125. Only four studies, two of them drawn from the broader project of which this Article is part, offer a data-driven accounting of qui tam litigation. See Engstrom, Harnessing, supra note 18, at 1288–1318 (using descriptive and regression approaches to explore repeat play among qui tam relators and their counsel); Engstrom, Public Regulation, supra note 85, at 1716–48 (using descriptive statistics and regression approaches to explore DOJ’s qui tam oversight); David Kwok, Evidence from the False Claims Act: Does Private Enforcement Attract Excessive Litigation?, 42 Pub. Cont. L.J. 225, 235–48 (2013) (offering descriptive statistics on repeat play among plaintiff’s counsel); David Kwok, Coordinated Private and Public Enforcement of Law: Deterrence Under Qui Tam 10–15 (Feb. 18, 2010) (unpublished manuscript), available at https://editorial express.com/cgi-bin/conference/download.cgi?db_name=ALEA2010&paper_id=375 (on file with the Columbia Law Review) (offering descriptive statistics on DOJ oversight). The remaining empirical studies limit their analyses to healthcare cases; treat only large, successful cases; or use only publicly available aggregated data. See Aaron S. Kesselheim et al., Whistle-Blowers’ Experiences in Fraud Litigation Against Pharmaceutical Companies, 362 New Eng. J. Med. 1832, 1834–36 (2010) (reporting relator-regret profiles who won large recoveries in healthcare cases); Aaron S. Kesselheim & David M. Studdert, Whistleblower-Initiated Enforcement Actions Against Health Care Fraud and Abuse in the United States, 1996 to 2005, 149 Annals Internal Med. 342, 343–48 (2008) (examining healthcare cases only); Broderick, supra note 116, at 963–67 (using aggregated data on filings and recoveries to explore regime workings). Advocacy groups have also published reports advancing empirical claims about the qui tam regime but suffering from serious methodological shortcomings. See, e.g., Nat’l Whistleblowers Ctr., Impact of Qui Tam Laws on Internal Corporate Compliance (2010), available at http://www.whistleblowers. org/storage/whistleblowers/documents/DoddFrank/nwcreporttosecfinal.pdf (on file with the Columbia Law Review) (analyzing 107 cases between 2007 and 2010, but considering only published judicial opinions). 126. Qui tam case information is maintained by the DOJ on its QTRACK internal tracking system. Note that roughly 3,000 qui tam suits remain under seal and are not part of the dataset, raising questions about the sample’s representativeness. These sealed cases likely fall into three categories: (i) filings from the past five years that remain under seal pending DOJ investigations; (ii) closed cases that remain under seal because they are subject to various privileges, including the state-secrets privilege (perhaps implicating national security); and (iii) other closed cases that remained sealed for other reasons, including simple neglect by the judge to unseal the case, accidental failure by the DOJ attorney to request unsealing upon case termination, or a successful relator effort to persuade the trial judge to keep the case sealed, typically because she remains employed by the defendant company. An interview with a former DOJ attorney suggests that the last type of case is likely to be concentrated in the period before 2000 or 2001 when the DOJ, under pressure from congressional overseers, changed its policy and began to take a more aggressive stance in advocating unsealing. See Interview with Former Attorney, Civil Fraud Section, U.S. Dep’t of Justice (Mar. 10, 2011) (on file with the Columbia Law Review). In 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1951 filings, identified only by their filing date. Second, electronic docket sheets were retrieved from PACER for each unsealed case from the relevant United States district court and information was extracted and merged with the DOJ-provided data regarding filing and termination dates, as well as the parties, law firms, and individual lawyers involved.127 The following discussion uses these unique data to begin the task of testing some of Part I’s theoretical propositions alongside persistent anecdotal claims about the qui tam regime. Part III first offers a detailed data-driven account of qui tam’s growth since 1986. Part IV continues the analysis by using a mix of further quantitative evidence and close-in case studies to test persistent claims that qui tam litigation has changed in its essential character over time and has not been subject to meaningful public control.

III. CHARTING QUI TAM’S RISE:ANATOMY OF A LITIGATION “EXPLOSION”

An initial assessment of the pathways down which the qui tam regime has traveled can begin by asking to what extent, and in what ways, qui tam litigation has grown since the FCA’s 1986 revival. To that end, this Part conducts a series of data-driven analyses of the post-1986 regime’s production and performance. Part III.A isolates and analyzes filing trends. Part III.B reports simple dollar-recovery trends. Part III.C then combines, and also reconfigures, the recovery analysis by reporting per-filing recovery rates and amounts over time as a way to gauge trends in the “efficiency” of qui tam suits at returning dollars to the federal fisc. The composite picture that emerges largely rejects widespread claims that qui tam litigation is in the midst of an inefficient “explosion” of enforcement effort or, tracking Part I’s broader concerns about manipulability of litigation flows, that qui tam enforcement activity has “lurched” from one extreme to another in response to a changing litigation environment. Rather, the data uniformly suggest a steady maturation of the regime.

A. Filings An obvious starting point in any evaluation of qui tam’s post-1986 growth is filing trends. Figure 3 enriches Figure 2’s summary portrait of the regime by providing a more detailed, annotated accounting of qui

sum, the sample used in this Article is likely unrepresentative in at least two respects: First, it contains more interventions than the full case population since 1986, and second, it contains more cases brought by former, not current, company employees. Possible impacts are discussed in connection with the presentation of specific empirical results below. 127. Docket sheet information was “scraped” using the “PERL” programming language. See generally The Perl Programming Language (PERL), http://www.perl.org/ (on file with the Columbia Law Review) (last visited Sept. 16, 2014). 1952 COLUMBIA LAW REVIEW [Vol. 114:1913 tam filings, plotted on a half-yearly basis, from 1986 to 2013.128 It shows that, far from a straight-line or accelerating increase, qui tam filings have ebbed and flowed over time. What best explains those trends? Are qui tam litigation flows a predictable response to discrete and measurable changes in the qui tam litigation environment? Or is there an extralegal and mostly unobservable story, perhaps relating to growing awareness and mobilization of potential plaintiffs, the emergence of an increasingly specialized qui tam plaintiffs’ bar, or growing speculation by opportun- istic actors within the system?

FIGURE 3: QUI TAM FILING TRENDS AND POSSIBLE EXPLANATIONS, 1986–2013

128. These filing counts include unsealed and sealed cases. Tallies of the latter came from a FOIA request asking that DOJ provide filing dates for sealed cases but no other identifying case information. 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1953

Any rigorous effort to answer such questions must account for a range of factors, each depicted in Figure 3, that might explain qui tam litigation flows over time. For instance, qui tam filings might be thought to track periods of partisan control of the executive if would-be relators perceive a partisan cast to the DOJ’s involvement in FCA cases and think DOJ involvement matters. Both possibilities enjoy support: Anecdotal evidence suggests the DOJ has been more supportive of qui tam cases during Democratic than Republican administrations,129 and DOJ intervention and ultimate recovery in cases are highly correlated.130 An additional trio of plausible factors, denoted in Figure 3’s top stack, include (i) idiosyncratically active relators, such as oil magnate Jack Grynberg, who filed a spate of roughly sixty separate qui tam suits in 1997 alleging that much of the oil and gas industry was engaged in a common practice of underpaying federal royalties;131 (ii) the terrorist attacks of September 11, 2001, which were associated with a precipitous decline in qui tam filings during the second half of that year; and (iii) the 2009 enactment of the Fraud Enforcement and Recovery Act

129. See, e.g., J. Randy Beck, The False Claims Act and the English Eradication of Qui Tam Legislation, 78 N.C. L. Rev. 539, 623 n.439 (2000) (highlighting congressional testimony by DOJ officials during first Bush Administration that large proportion of qui tam cases lacked merit). Compare Common Legislative Encroachments on Executive Branch Authority, 13 Op. O.L.C. 248, 250–51 (1989) (arguing, during first Bush Administration, DOJ should argue FCA’s qui tam provision is unconstitutional), with The Constitutional Separation of Powers between the President and Congress, 20 Op. O.L.C. 124, 142 n.52 (1996) (disavowing, during Clinton Administration, holdings of earlier Appointment Clause opinions, including challenge to qui tam constitutionality). In addition, the George W. Bush Administration drew allegations that its DOJ sat on qui tam cases relating to controversial war efforts in Iraq and Afghanistan. See Engstrom, Public Regulation, supra note 85, at 1714–15, 1745 (collecting source materials supporting these views). By contrast, the Clinton DOJ inaugurated a high-profile antifraud effort in the healthcare arena, and the Obama DOJ has repeatedly affirmed its intention to use the FCA aggressively. See Civil Division of the United States Department of Justice: Hearing Before the Subcomm. on Commercial & Admin. Law of the H. Comm. on the Judiciary, 111th Cong. 14–15 (2010) (statement of Tony West, Assistant Att’y Gen. of the United States) (expressing strong interest of Obama DOJ in expanding FCA enforcement activity); Jaeger & Diesenhaus, supra note 9, at S32 (noting Clinton DOJ’s “high priority” for combating health fraud). Note here that Figure 2’s trendlines run contrary to other, competing theories of the effect of partisan control of the executive on litigation flows. For instance, an activist agency could crowd out private enforcement efforts by moving against the most attractive enforcement targets and leaving private enforcers with a pool of more marginal litigation opportunities. See, e.g., Christian Langpap & Jay P. Shimshack, Private Citizen Suits and Public Enforcement: Substitutes or Complements?, 59 J. Envtl. Econ. & Mgmt. 235, 236 (2010) (considering possibility of crowding out private enforcement in environmental enforcement). 130. See infra notes 240–241 and accompanying text (noting high correlation between DOJ intervention and ultimate recovery success). 131. Data obtained from the DOJ reflect that Grynberg filed sixty separate suits in 1997. Though all remain under seal, the filing dates and parties can be identified by examining the data. See, e.g., United States ex rel. Grynberg v. Exxon Pipeline Co., No. 97-198B (D. Wyo. filed Aug. 6, 1997). 1954 COLUMBIA LAW REVIEW [Vol. 114:1913

(FERA), which made qui tam suits easier to bring by expanding the scope of FCA liability and removing certain procedural hurdles.132 Two final factors, depicted in Figure 3’s lower stacks, likewise seem critical to any rigorous accounting of qui tam filing trends. First, the period since 1986 has seen substantial growth in federal healthcare and defense-procurement spending. More concretely, health and defense- procurement expenditures—which can be thought of as constituting the bulk of “fraud-eligible” federal funds from a qui tam enforcement perspective—more than doubled in real-dollar terms over the study period and saw particularly rapid growth throughout the 2000s, due at least in part to a new prescription-drug entitlement under Medicare and—though it is not as readily discerned from Figure 3—a pronounced post-9/11 rise in defense spending.133 As one qui tam advocate has put it, the “swamp” of fraud opportunities has grown substantially over the life of the post-1986 regime.134 Qui tam filings, it follows, might merely track the size of the swamp. Second, the health of the economy has varied drastically since 1986, as demonstrated by Figure 3’s plot of unemployment rates over the study period, with slumps in 1992 and in 2002 to 2004 followed by the financial

132. See Fraud Enforcement and Recovery Act of 2009 (FERA), Pub. L. No. 111-21, § 4(a), 123 Stat. 1617, 1621–23 (codified as amended at 31 U.S.C. § 3729 (2012)) (clarifying provisions of FCA). Among other things, FERA overruled a 2008 Supreme Court decision, Allison Engine Co. v. United States ex rel. Sanders, 553 U.S. 662, 671–72 (2008), insulating subcontractors from qui tam suits. See Matthew Titolo, Retroactivity and the Fraud Enforcement and Recovery Act of 2009, 86 Ind. L.J. 257, 260 (“FERA ‘legislatively overrules’ Allison Engine Co.....”). FERA also reduced the FCA’s intent requirement, FERA § 4(a); expanded FCA conspiracy liability, id.; increased whistleblower protections, id.; and beefed up the DOJ’s investigatory powers, id. § 4(b). Note, however, that FERA may not be a good explanation for the immediate observed spike in filings because it did not have retroactive effect and so did not cover claims for payment predating enactment. As a result, one would expect its impact to lag, with many cases based on 2009 claims for payment not filed until 2010 or later. 133. See Office of Mgmt. & Budget, Table 16.1—Outlays for Health Programs: 1962– 2019, available at http://www.whitehouse.gov/sites/default/files/omb/budget/fy2015/ assets/hist16z1.xls (on file with the Columbia Law Review) (last visited Oct. 7, 2014) (providing consolidated health-program-spending data); Office of Mgmt. & Budget, Table 3.2—Outlays by Function and Subfunction: 1962–2019, available at http://www.white house.gov/sites/default/files/omb/budget/fy2015/assets/hist03z2.xls (on file with the Columbia Law Review) (last visited Oct. 7, 2014) (summarizing defense-procurement spending). In addition to health and defense-procurement outlays, it is noteworthy that overall procurement spending—often referred to in budget-speak as “Federal Contract Actions”—has more than doubled since 1986. See U.S. Census Bureau, U.S. Dep’t of Commerce, CFFR/10, Consolidated Federal Funds Report for Fiscal Year 2010, at xx fig.4 (2011), available at http://www.census.gov/prod/2011pubs/cffr-10.pdf (on file with the Columbia Law Review) (showing increase in procurement spending by over $200 billion dollars since 1983). 134. See Pharma Qui Tam Caseload Jumps to 180, Rx Compliance Rep., Jan. 16, 2007, at 1, 8, available at http://rxcompliancereport.com/issues/2007/RxComp0116.pdf (on file with the Columbia Law Review) (referencing earlier statement by qui tam advocate Patrick Burns). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1955 meltdown beginning in 2008.135 These macroeconomic ebbs and flows are potentially important, though the predicted direction of the effect is ambiguous. On the one hand, qui tam filings may rise during recessions, when the pool of unemployed relators with information about fraud, ill will toward former employers, and little to lose expands relative to more prosperous times.136 On the other hand, a soft labor market could muzzle would-be whistleblowers who remain company employees by raising the potential cost of termination or other forms of retaliation.137 Similarly, finance scholars have suggested that corporate wrongdoing is cyclical, increasing during prosperous times when it is more easily hidden and lucrative or, alternatively, peaking at the beginning of economic down- swings when management bends rules to prop up flagging profits.138 Unfortunately, statistical measurement of time-series relationships of this sort is famously fraught. This is particularly true where, as is the case here, there are multiple parallel growth trends, creating a substantial risk of “spurious” regression.139 As a concrete example, a simple regression of qui tam filing counts on healthcare spending would likely find a substantial relationship, since both variables exhibit a general upward trend across the study window. But such a finding would not necessarily support an inference that healthcare spending caused the parallel growth

135. Figure 3’s unemployment figures are derived from a data series titled “Unemployment Rate (Seasonally Adjusted) - LNS14000000” and maintained by the United States Bureau of Labor Statistics. See Top Picks, U.S. Bureau of Labor Statistics, http://data.bls.gov/cgi-bin/surveymost?bls (on file with the Columbia Law Review) (last visited Oct. 14, 2014) (select “Unemployment Rate (Seasonally Adjusted) - LNS14000000” from “Employment” list; then select “Retrieve Data” for HTML output; then select “Download: xlsx” for Excel output). 136. See William E. Kovacic, The Civil False Claims Act as a Deterrent to Participation in Government Procurement Markets, 6 Sup. Ct. Econ. Rev. 201, 233 (1998) [hereinafter Kovacic, Deterrent] (“Layoffs have created a large pool of potential relators who have less to risk by way of damaging a relationship with an existing employer, and who may have fewer inhibitions with respect to experimenting with new theories of CFCA liability.”). 137. See Farhang, Litigation State, supra note 15, at 22 (using expected monetary gains and losses to predict prospective plaintiff decision to litigate). The basic logic here is that hard economic times may make whistleblowing too risky for employees who may not be able to find alternative employment if terminated in retaliation. 138. See, e.g., Paul Povel, Rajdeep Singh & Andrew Winton, Booms, Busts, and Fraud, 20 Rev. Fin. Stud. 1219, 1222 (2007) (discussing correlation between business cycle and fraud). 139. For a classic account of why statistical analysis often finds spurious correlation between trending time series, see C. W. J. Granger & P. Newbold, Spurious Regressions in Econometrics, 2 J. Econometrics 111, 117–19 (1974). More technically, where two or more sets of time-series data points are nonstationary—that is, do not follow a regular path around a mean—their error terms are liable to autocorrelate, producing spurious regression results. 1956 COLUMBIA LAW REVIEW [Vol. 114:1913 in filings. Rather, the two measures could be drifting in similar directions for entirely independent, causally unrelated reasons.140 Still, Figure 3’s descriptive presentation permits one to venture a pair of broad conclusions that cut against widespread characterizations of qui tam’s post-1986 growth as a litigation “explosion.” The first comes from consideration of filing counts alone: While qui tam filings grew rapidly at the dawn of the modern regime, and also appear to be in the midst of a substantial uptick, they arguably plateaued around 400 lawsuits per year in the intervening period, between roughly 1997 and 2008. Of course, the recent uptick makes it possible that the regime has shifted into a new phase: A key question, unanswerable with only thirty years of data, is whether the current clip of 600 qui tam filings per year is the new normal, or whether filings will revert to the 400-case steady state that had prevailed during much of the post-1986 regime. In addition, the filing data cannot fully reject the possibility of lurching behavior of the sort Part II.A theorized, given the many ebbs and flows in filings over time. But it is simply not the case that qui tam filings have experienced accelerating growth or something one could call a litigation “explosion.” Rather, the most apt way to describe the post-1986 growth of the system is steady maturation. The second broad conclusion derives from Figure 3’s consideration of filing trends alongside possible explanations for those trends: The rise in health spending, defense spending, and unemployment rates plausibly account for a substantial amount of the rise in qui tam filings, leaving less room for alternative explanations. More specifically, the significant long- run growth in healthcare spending offers a plausible explanation for some of the growth in qui tam filings, particularly given the dominance of healthcare claims within the system. Similarly, the post-9/11 rise in defense spending and the more recent dramatic spike in unemployment as a result of the 2008 downturn both map well to observed growth in qui tam enforcement activity. Importantly, once these possible relationships are accounted for, the residual, underlying growth in qui tam filings that

140. There are two ways one might mitigate spurious regression concerns and explore the relationship between filing counts and Figure 3’s covariates. One is to take first-differences of all numeric variables of interest and then run a regression model. By detrending the variables in this manner, one can estimate the relationship between movement in filings and key variables (e.g., whether filing counts have increased more quickly when health spending increased more quickly). This approach, however, did not find any statistically significant relationships. Another option is to examine the relationship between filings and covariates using a Generalized Additive Model (GAM), where each of the component functions is estimated using a smoothing spline. This modeling approach did not find a positive effect of time on the count of cases filed, though it did find a possible relationship between health spending and filings. These findings support the conclusion that, once a range of intuitive explanations for filing growth are taken into account, qui tam litigation cannot fairly be characterized as in the midst of a filing “explosion.” The results of both of the above analyses are available upon request to the author and on file with the Columbia Law Review. 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1957 can be attributed to other potential causes—whether wasteful opportunism, growing awareness of the regime among would-be relators, or growing capacity of the relators’ bar to identify whistleblowers with actionable information—may well be quite small. Here again, Figure 3’s accounting, while plainly falling short of decisive statistical proof, tends to undermine widespread characterizations of a qui tam “explosion.”

B. Recoveries Recovery amounts offer a second way to evaluate qui tam’s post-1986 growth. Figure 2 shows that recoveries have climbed more or less in tandem with filings, reaching a recent annual peak of more than $3 billion. But questions remain: How, if at all, has the distribution of qui tam recoveries changed over time? Have some flavors of qui tam claims grown faster than others? Figure 4 offers an initial cut at answering such questions.141 The top stack presents a time-series scatterplot of qui tam recoveries from 1986 through the first half of 2013, with recoveries above $100 million indi- cated by an industry-specific letter designation. The bottom stack plots mean and median recovery amounts over that same time period. To be sure, the scatterplot reveals some outsized recoveries, with markedly larger recoveries appearing with increasing regularity after roughly 1998. The scatterplot likewise shows the dominance of healthcare cases (denoted with an “H”) at the high end of recovery dollar amounts. Indeed, of the sixty-four qui tam recoveries above $100 million, nearly eighty percent have been healthcare cases as opposed to defense-procurement (“D”) cases or other, nonhealth and nondefense (“O”) cases.

141. All dollar amounts in Figure 4 are expressed in 2013 dollars. Annual means and medians are calculated based on calendar year and presented as Lowess curves. Note two measurement issues raised by Figure 4’s defendant-focused analysis. First, the analysis treats as a single data point settlements reached on the same day against multiple but legally related defendants. As an example, a $50 million settlement against Lockheed Martin Space Systems, Inc., and a $60 million settlement that same day in the same lawsuit against Lockheed Martin Service, Inc. is treated as a single recovery. Second, Figure 4’s analysis treats as a single data point settlements resolving multiple separate qui tam actions brought by distinct relators. An example is the Pfizer case, which the DOJ-provided data say yielded a settlement on October 16, 2009, totaling $677 million (in 2009 dollars), and resolved five separate cases filed by five distinct relators. The key point here is that such settlements were brokered by the DOJ, with settlement amounts then allocated to each relator for purposes of determining relator shares. But from a system-production perspective—and, in particular, an inquiry focused on gauging growth in recoveries over time—the most accurate measure is to determine whether the system is generating increased liability from the perspective of regulated parties. Part III.C turns to a critically different analysis focused on trends in the “efficiency” of individual-relator-filed actions in returning funds to the federal fisc. 1958 COLUMBIA LAW REVIEW [Vol. 114:1913

FIGURE 4: QUI TAM RECOVERIES BY CASE TYPE AND CLOSURE DATE, JANUARY 1989–JUNE 2013

At the same time, however, Figure 4 suggests something less than a qui tam “explosion.” Significantly, while the lower stack shows that mean recoveries have roughly doubled since 2000 (and nearly quadrupled since 1990), the upper stack’s scatterplot suggests that a substantial portion of this growth is attributable to a handful of especially large settlements of roughly $500 million or more (in 2013 dollars), most of them against pharmaceutical companies Abbott Laboratories ($582 million in 2012), Eli Lilly ($480 million in 2009), GlaxoSmithKline ($1.53 billion in 2012 and $471 million in 2010), and Pfizer ($750 million in 2009).142 When these outliers are dropped from the sample, mean recoveries have in fact stayed roughly constant since the early 2000s. A similar conclusion follows from consideration of median recovery amounts, which have hovered around $1 million throughout the life of the post-1986 regime, with the exception of a slight uptick to $2 to 3 million during the 2010 to 2012 period. This once more underscores the top-heavy nature of the qui tam regime’s recent growth, as the median calculation gives less weight to the larger recoveries that appear in the top stack of Figure 4. It also bolsters the conclusion that, apart from a handful of outsized recoveries, the overall story in terms of recovery dollar amounts seems to be relative stability, not accelerating growth, over the past decade.

142. A sixth large recent settlement, with the State and City of New York for $585 million in 2009, also resolved claims involving healthcare fraud but was not directed at a pharmaceutical company. 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1959

FIGURE 5: PROPORTION OF TOTAL RECOVERY COUNTS AND RECOVERY DOLLARS IN QUI TAM CASES BY RECOVERY SIZE OVER TIME, 1989–2012

Figure 5’s barplot offers further confirmation of this point by summarizing the proportion of total recovery counts (the left-hand barplot) and total recovery dollars (the right-hand barplot) accounted for by different-sized recoveries (i.e., less than $1 million, $1 to 10 million, $10 to 100 million, and more than $100 million) as tracked across four- and five-year time windows between 1989 and 2008 (i.e., 1989 to 1993, 1994 to 1998, 1999 to 2003, 2004 to 2008, and 2009 to 2012). The analysis confirms that the qui tam regime has been characterized by substantial high-end growth. Indeed, $100-million-plus cases have grown from one percent of successful case counts to three percent. More importantly, $100-million-plus recoveries now account for more than sixty percent of total recoveries in dollar terms, up from less than twenty-five percent in the regime’s early years, and up roughly five percent from the regime’s middle period of development. Similarly, smaller recoveries, including those falling into the less-than-$1-million and less-than-$10-million categories, show some decline, both as a proportion of successful-case counts and as a proportion of recovery dollars. Here again, however, the overall portrait, while not exactly static, does not suggest an accelerating increase apart from a few high-end recoveries.143

C. Per-Filing “Efficiency” If, as Part III.A demonstrates, qui tam filings may not exhibit a strong underlying upward trend after controlling for observable features of the litigation environment, and if, as Part III.B demonstrates, growth in recovery amounts appear attributable to a small number of outsized

143. Also potentially noteworthy is the decline of the proportion of recoveries that fall below the $1 million threshold. One interpretation is that the regime is generating fewer nuisance settlements, which would also be inconsistent with a qui tam “explosion.” For a similar analysis, see James D. Cox & Randall S. Thomas with Dana Kiku, SEC Enforcement Heuristics: An Empirical Inquiry, 53 Duke L.J. 737, 770 (2003) (examining low-value settlements in securities suits as measure of number of frivolous suits). 1960 COLUMBIA LAW REVIEW [Vol. 114:1913 settlements, is there nonetheless evidence of a “gold rush” of qui tam enforcement activity?

FIGURE 6: QUI TAM CASE OUTCOMES BY FILING YEAR, JANUARY 1988–JUNE 2013

Figure 6 sheds light on this question via an analysis of qui tam cases filed since 1986 that differs from Figure 4 in two critical ways. First, while Figure 4 considered only winning cases in calculating mean and median recoveries, Figure 6 calculates success rates while including all closed cases, win or lose, in the denominator. The analysis thus captures a version of what one might dub enforcement “efficiency”—that is, the average success rate or dollar return to the federal fisc per qui tam case filed. Second, Figure 6 tallies and reports outcomes keyed to case filing date, rather than termination or settlement date. This approach treats cases filed in a given year as cohorts moving through the system, reveal- ing differences between cases filed at distinct stages in the regime’s 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1961 life.144 If the increase in high-end recoveries beginning in the later 1990s—and the press coverage accompanying them—has prompted a “gold rush” of enforcement activity, then one might expect to see declining recovery rates, declining per-filing returns to the federal fisc, or both, as relators and their attorneys file more (and more marginal) cases seeking large payouts. At first glance, Figure 6’s various stacks suggest declining relator fortunes since 1986. More specifically, the top stack’s accounting of total recovery dollars by filing year suggests that recoveries grew substantially among cases filed in the middle period of the post-1986 regime but have declined since. The middle stack’s plot of annual recovery rates (i.e., the proportion of cases filed in a given year that ultimately produced a positive-dollar recovery) tells an even starker story, suggesting that success rates have suffered a steady decline over the life of the post-1986 regime, falling from nearly fifty percent in the early years of the regime (1987 to 1995) to roughly thirty-three percent in the middle years (1996 to 2003) then dipping still further from roughly twenty percent to zero in more recent years (2004 to 2013). The bottom stack, which combines the analyses presented in the first two stacks and thus captures the regimes’ per-filing “efficiency” at returning funds to the federal fisc, once more shows steady decline. This is particularly evident from the dotted line, which plots the results of a local (Lowess) regression on annual mean per-filing recoveries from 1986 to 2013 and implies that per-filing returns to the federal fisc began around $10 million at the dawn of the regime but have since steadily declined to nearly zero.145 However, it is critically important to note a pair of measurement challenges in any analysis of qui tam case outcomes pegged, as in Figure 6, to filing years. First, successful qui tam cases have historically had gestation periods that are substantially longer than unsuccessful ones, both in drawing a DOJ intervention decision (and thus coming out from under court seal) and in reaching final judgment thereafter.146 The

144. Note a further way in which the analysis differs from Figure 4’s: While Figure 4’s defendant-focused analysis collapses together multiple cases achieving a single omnibus settlement against a common defendant, Figure 6 is plaintiff-focused, collapsing multiple lawsuits into a single “action” brought by a common relator and/or counsel close in time to one another. More specifically, the analysis collapses qui tam cases with common counsel or a common relator, wherein two of the following three conditions obtained: The cases were filed, drew a DOJ-election decision, or were settled or terminated in a common quarter of the calendar year. Doing so controls for the fact that some relators file “omnibus” actions naming multiple defendants in a single lawsuit while other relators file numerous serial lawsuits, whether in the same district court or across district courts. For broader discussion of this coding approach, see Engstrom, Harnessing, supra note 18, at 1290–91 (explaining advantages and disadvantages of this data-trimming method). 145. Note that these results hold when the sample is constrained to only healthcare cases—the only case type for which there are enough data to reliably detect trends. 146. As an example, cases that closed in 2011 and produced recoveries were under DOJ investigation for an average of 1,484 days, while cases that did not produce recoveries were under DOJ investigation for an average of 524 days. Similarly, successful cases took 1962 COLUMBIA LAW REVIEW [Vol. 114:1913 result is a right-censoring problem: Because losing cases from the more recent filing years are systematically more likely than winning cases to have reached a DOJ case-election decision and be out from under seal, the data sample for these more recent years will be over populated with losing cases and under populated with successful cases.147 Figure 6’s results thus likely underestimate mean and median per-filing recoveries after roughly 2005. Indeed, after 2008, cases that remain under DOJ investigation and thus under seal outnumber unsealed cases, making inferences particularly unsound.148 A second consideration is that sealed qui tam suits filed during more recent periods are more likely to be successful than sealed qui tam suits filed during earlier periods.149 For recent filing years, the most likely reason a case remains sealed (and is thus excluded from Figure 6’s analysis) is that it awaits a DOJ election decision. Some of these cases will ultimately achieve a recovery. Not so for cases filed earlier in the life of the regime, most of which have long since terminated but remain sealed because the relator voluntarily dismissed the complaint in response to a DOJ declination and then persuaded the court not to lift the seal in an effort to protect her anonymity.150 Thus, while Figure 6’s results likely underestimate per-filing recoveries for recent filing years, they also likely overestimate per-filing recoveries during earlier stages of the regime by excluding nontrivial numbers of unsuccessful but sealed cases. Given data shortcomings, perhaps the best approach is to discount or simply ignore the results after 2008, when the number of sealed cases begins to outstrip the number of unsealed cases. The second solid an average of 1,379 days following the DOJ’s case-election decision to reach closure, while unsuccessful cases took only 297 days. Part IV.C infra examines case gestation periods in more detail. 147. In addition, excluded successful cases may be large relative to successful cases in the dataset. One such example is the $2.2 billion settlement with J&J, which opened this Article. The settlement resulted from cases filed in and around 2008 but settled beyond the time window covered by the data. See supra notes 1–3 and accompanying text. 148. For instance, the DOJ-provided data report that 105 of 379 qui tam filings in 2008 remain under seal, or roughly twenty-eight percent of the total. For 2012, 587 of 647 remain under seal, or ninety-one percent. 149. The qui tam regime generated some 8,000 lawsuits between 1986 and 2012, and more than 5,000 of these have emerged from under seal and are included in the data sample. That leaves some 2,800 cases that remain sealed and are unobservable, including roughly twenty cases per year from 1986 to 1995, another 100 cases per year between 1995 and 2007 as the regime increased in volume, and 600 cases per year by 2012. See supra Part III.A (describing qui tam lawsuit filing trends); see also supra note 6 (citing statistics on qui tam filings and recoveries). 150. For example, relators may seek to protect their anonymity where they remain the defendant’s employee or are seeking employment elsewhere within the industry. See Interview with Former Attorney, Civil Fraud Section, U.S. Dep’t of Justice, in Washington, D.C. (Mar. 9, 2011) (on file with the Columbia Law Review) (providing examples of relators depending on defendant for livelihood); see also Email from Former Attorney, Civil Fraud Section, U.S. Dep’t of Justice (Oct. 2, 2014) (on file with the Columbia Law Review). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1963 regression line in Figure 6 does just that, reestimating the per-filing recovery trend up through 2008. Constraining the analysis in this way changes the interpretation substantially, implying that relator fortunes have declined only slightly. It is even possible that relator fortunes have not declined at all, given the likely inflation of per-filing recoveries prior to 2000 because of the exclusion of sealed but unsuccessful cases from consideration. To be sure, it remains possible that a fuller case accounting covering the years from 2009 to 2012 would reveal a convulsive shift in relator fortunes once all cases have emerged from under seal and reached final resolution. For now, however, the best conclusion is that, contrary to heated claims about a “gold rush” of qui tam enforcement efforts, the system has not grown substantially less “efficient” over time in returning funds to the federal fisc.

** *

In sum, the composite evidence points decisively away from widespread claims that qui tam enforcement efforts are in the midst of an inefficient “explosion.” Still, it is important to emphasize the limitations of the above analyses. Some of these have already been noted, including the right censoring of the data for analyses pegged to filing years. More broadly, the collected analyses plainly fall short of a bottom- line social-welfare judgment as to the net benefit or cost of qui tam litigation. As many commentators have noted, such judgments are particularly fraught regarding litigation regimes, and qui tam is no exception.151 Finally, and most important of all, while the above analyses offer an initial, data-based survey of the qui tam enforcement landscape, they largely abstract from key questions about whether possible changes in the character and nature of qui tam enforcement efforts have moved the elaboration of the FCA’s antifraud mandate in directions that have evaded meaningful public control. The next Part turns to these subtly different issues.

151. On social-welfare assessments of litigation regimes, see supra note 14 and accompanying text (discussing social-welfare implications of public–private enforcer choice); see also Hazel Genn, Judging Civil Justice 181 (2010) (“[T]he machinery of civil justice sustains social stability and economic growth by providing public processes for resolving civil disputes, for enforcing legal rights and for protecting private and personal rights.”). A welfarist assessment of the FCA would necessitate quantification of numerous items on either side of the cost–benefit ledger, from the amount of fraud deterred to total litigation costs and even, according to one commentator, the cost of reduced competition in government-contractor markets thinned by the threat of FCA liability. On the latter, see Kovacic, Deterrent, supra note 136, at 239 (noting the efficiency cost of thinned contractor markets). 1964 COLUMBIA LAW REVIEW [Vol. 114:1913

IV. QUI TAM’S CHANGING NATURE,PRIVATE LEGAL INNOVATION, AND THE PUZZLE OF PUBLIC CONTROL This Article has thus far considered a large body of theory regarding private enforcement’s pathways and tested some initial hypotheses by exploring key dimensions of the FCA regime’s post-1986 growth. Having laid this theoretical and empirical foundation, this Article now turns to subtler questions about private legal innovation and public control by assessing widespread claims that qui tam enforcement has changed in its essential character since 1986 and, further, has driven elaboration of the FCA’s antifraud mandate in politically unaccountable directions. To be sure, some of the analysis that follows quickly runs up against methodological and conceptual challenges. For starters, the counterfactual nature of portions of the inquiry makes direct empirical testing difficult. Above all, how can one know with certainty that qui tam enforcement efforts have steered elaboration of the FCA’s antifraud mandate in directions it would not have traveled in an imagined world of purely public enforcement? More concretely, does the absence of congressional or agency action regarding a particular theory of FCA liability suggest legislative or administrative acquiescence or even support? Or does it instead suggest a lack of institutional will or capacity to wade in at all? While acknowledging such complexities, this Part deploys three distinct empirical strategies in an effort to explore the political and democratic control issues raised in Part I’s theoretical work-up and, more specifically, whether the presence of a qui tam enforcement mechanism has driven the FCA down interpretive pathways it would not have traveled in a world of purely public enforcement. Part IV.A offers evidence on the changing character of qui tam litigation by demonstrating a pronounced upward trend in the mean number of defendants per action during the first half of the post-1986 FCA regime. These findings lend credence to, but cannot directly prove, anecdotal claims that qui tam enforcement efforts have increased in regulatory scale and scope over time by pushing into regulatory interstices, where political control is apt to be less dependable. Part IV.B offers a more granular perspective than has been attempted to this point by isolating several distinct FCA enforcement efforts and tracing their origins and evolution over time as a way to generate some more-general insights about the nature of private legal innovation and public control within the system. Part IV.C closes the analysis by reporting basic empirical data that lay bare key trends in the DOJ’s exercise of its case-level gatekeeper powers under the FCA as a final gauge of the degree of public control within the system.

A. Litigation Scale and Scope The empirical results presented in Part III broadly showed that qui tam’s post-1986 growth has been more steady maturation than explosion. But are there other ways in which the regime has changed over time? 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1965

Critics think so. As noted previously, an influential line of criticism says that qui tam suits have increasingly moved away from targeting uncontro- versial, “raw” frauds by lone individuals or entities—e.g., simple “upcoding” of charges in seeking government reimbursement for healthcare services—and toward more arguable, gray-area misconduct engaged in by much or all of an industry.152 Many of these same critics also warn of rising use of “false certification” claims focused on relatively minor regulatory infractions. As the more critical account set forth in Part I suggests, the rising scale and scope of private enforcement efforts will necessarily transfer lawmaking and agenda-setting power from public to private actors, thereby undermining democratic control. To what extent does the evidence support these concerns? As an initial matter, there is substantial anecdotal evidence that certification claims now form a nontrivial part of qui tam enforcement activity. Some of the best evidence is a long line of appellate decisions confronting certification claims and imposing a “materiality” standard—partially codified by the 2009 FERA amendments—requiring a relator to show that a defendant’s alleged falsehoods mattered to the government’s decision to pay on a contract.153 The fact that materiality arises most squarely in certification cases fuels an inference that such cases are, at the least, modestly prevalent and further suggests, though plainly cannot prove, that qui tam enforcement efforts have moved into regulatory gray areas. Substantial anecdotal evidence likewise supports the claim that qui tam actions have increasingly taken multidefendant, pan-industry forms. A canvass of the data reveals numerous recent examples of large-scale, pan-industry suits that make the several dozen qui tam suits filed by oil executive Jack Grynberg against much of the oil and gas industry in 1997—as described previously in Part III.A’s filing analysis154—look small by comparison. An extreme example is a 4,000-defendant colossus filed in Connecticut district court in 2002 alleging that seemingly every hospital in the region improperly billed the Medicare and Medicaid

152. See supra note 9 and accompanying text (describing lawsuits tethering FCA’s open-ended fraud mandate to statutes without private rights of action); supra note 118 and accompanying text (discussing lawsuits targeting violations of ambiguous regulatory mandates, and compiling critiques of qui tam suits). 153. See Holt & Klass, supra note 120, at 8 n.53 (collecting case law imposing materiality requirement); see also Fraud Enforcement and Recovery Act of 2009 (FERA), Pub. L. No. 111-21, § 4(a)(1), 123 Stat. 1617, 1621 (codified as amended at 31 U.S.C. § 3729(a)(1)(B) (2012)) (imposing liability on “any person who . . . knowingly makes, uses, or causes to be made or used, a false record or statement material to a false or fraudulent claim” (emphasis added)). Note that the added “materiality” requirement in the FERA amendments applied only to claims under § 3729(a)(1)(B) and (G), both of which concern a “false record or statement” made in connection with a false claim, not claims under § 3729(a)(1)(A), the more general false claims provision. § 3729(a)(1)(A), (B), (G). 154. See supra note 131 and accompanying text (discussing Jack Grynberg’s sixty qui tam suits in 1997 against oil and gas industry). 1966 COLUMBIA LAW REVIEW [Vol. 114:1913 programs for radiology services.155 Other, more representative examples include a 2002 qui tam suit in which a consortium of specialized relator- side firms filed a qui tam suit against 61 named and 200 unnamed drug manufacturers, labelers, distributors, and retailers, alleging that each had facilitated false claims by healthcare providers participating in the Medicare and Medicaid programs by actively marketing drugs for uses that had not yet won FDA approval.156

FIGURE 7: LOG NUMBER OF DEFENDANTS PER QUI TAM “ACTION,” 1988–2013

The question remains, however, whether such cases are idiosyncratic or whether they have generated measurable shifts in the nature of qui tam enforcement over time. Figure 7 offers evidence on this question via a scatterplot of the logged number of defendants in a random sample of 1,000 qui tam “actions”157 filed since 1986. It then fits a regression line (and ninety-five percent confidence band) implying a roughly two-fold increase in the mean number of defendants per qui tam “action”—in

155. United States ex rel. Smith v. Bridgeport Hosp., No. 3:02-cv-01205-PCD (D. Conn. filed July 12, 2002) (showing, on docket sheet, original lawsuit filed against 4,000 defendants before complaint was amended). 156. Complaint at 10–11, United States ex rel. Conrad v. Wyeth Biopharma, Inc., No. 1:02-cv-11738-RWZ (D. Mass. Aug. 29, 2002). Another recent example is an omnibus complaint—filed in 2007 by the same firms as the 2002 case and, interestingly, what appears to be the spouse of the relator from the earlier case—accusing forty-eight pharmaceutical companies of misreporting their drugs as “Non-Innovator” drugs to avoid paying hundreds of millions of dollars in rebates via the Medicaid Rebate Program. Second Amended Complaint at 3–4, United States ex rel. Conrad v. Abbott Labs., No. 1:07-cv-03176-RDB (D. Md. July 1, 2009). 157. An “action” here is defined the same way as in Figure 4’s analysis. See supra note 141 (describing coding of single “actions”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1967 unlogged terms, from an average of 1.5 defendants per action to more than three defendants per action—during the first half of the post-1986 regime. This lends substantial credence to claims that qui tam suits have increased in scale and scope since the dawn of the current regime. Still, interpretive caution is clearly in order. Because the results are pegged to filing years, the analysis is once more dogged by a right- censoring problem after roughly 2008. Exogenous factors, such as con- solidation or fragmentation of various sectors of the healthcare industry, rather than shifts in the nature of qui tam litigation, might be driving the results as well. And even if exogenous factors of this sort are not behind the observed rise in multidefendant suits, Figure 7 permits few conclusions about whether the trend is a good or bad thing. In particular, it reveals little about whether the increase in multidefendant suits reflects simple growth in the relator bar’s enforcement capacity, rising use of statutory and regulatory ambiguities to target larger swathes of industry, or a trend toward “blunderbuss” filings that are in fact no more effective, or even less effective, at returning funds to the federal fisc. One would need to perform a more fine-grained analysis that links Figure 7’s defendant counts to claim types and defendant-level success rates to adjudicate between these possibilities. A final interpretive challenge concerns the apparent leveling off of multidefendant actions beginning around 2000. One plausible explanation that enjoys at least some anecdotal support is judicial push back. One can glimpse judicial exasperation in recent qui tam decisions, including a 2011 decision in which the Fifth Circuit, in affirming dismissal of a suit brought against “the entire Mississippi [durable medical equipment] industry,” tartly noted: “It takes no particular knowledge or effort to describe a general scheme of fraud and then list arbitrarily a large group of possible perpetrators.”158 Judicial vexation may underpin a range of efforts by lower courts to fashion doctrinal structures that limit pan-industry qui tam actions. Such efforts include increasing judicial openness to dismissal on the basis of Rule 9(b)’s heightened pleading standard for fraud claims;159 narrow construal of the FCA’s “public disclosure” bar to preclude qui tam suits based on public reports of industry- wide fraud, even against industry actors not specifically named in those reports;160 and judicial creation of a “materiality” standard, partially codi-

158. United States ex rel. Jamison v. McKesson Corp., 649 F.3d 322, 331 & n.17 (5th Cir. 2011). 159. Fed. R. Civ. P. 9(b); see also Elameto, supra note 8, at 821–24 (collecting cases regarding Rule 9(b)’s important gatekeeper function). 160. Courts have repeatedly struggled with this question. For a sampling, see United States ex rel. Baltazar v. Warden, 635 F.3d 866, 867–68 (7th Cir. 2011) (holding public disclosure indicating fifty-seven percent of Medicare claims from chiropractors are fraudulent was insufficient to bar suit against one chiropractor because it still “takes a provider-by-provider investigation to locate the wrongdoers”); United States ex rel. Fine v. Sandia Corp., 70 F.3d 568, 571 (10th Cir. 1995) (holding disclosure that two of nine laboratories were engaging in fraudulent practice “sufficiently alerted the government to 1968 COLUMBIA LAW REVIEW [Vol. 114:1913 fied by Congress in the 2009 FERA amendments, requiring a showing that the government would not have paid on a claim had it known of the breach or statutory or regulatory violation, thus stymieing certification claims in particular.161 As a result, the evidence presented in Figure 7 may well support concerns that private enforcement, and qui tam enforcement in particular, will exhibit increasing regulatory scale and scope, particularly early in the life of a regime. At the same time, however, it may also suggest that, contrary to Part I’s concern about courts’ limited will and capacity, judges are eminently capable of policing such concerns.

B. Case Studies in Private Legal Innovation and Public Control Evidence of a rise in multidefendant actions lends credence to the claim that qui tam litigation has increased in scale and scope since 1986. This evidence likewise suggests, though it cannot prove, that qui tam enforcement efforts have moved into legal interstices, where political and democratic control may be less dependable. But the empirical analysis to this point has mostly sidestepped a pair of critical claims about private enforcement sketched in Part I’s theoretical workup: Private enforcers will progressively target regulatory ambiguities left by legislative or administrative inertia, and piecemeal litigation efforts will thus push legal mandates down interpretive pathways they would not travel with purely public enforcement. Progress testing these claims can only be achieved, if at all, through a more granular assessment of actual litigation efforts. What follows are case studies profiling three distinct “flavors” of qui tam litigation that account for a nontrivial proportion of total FCA recoveries. Though plainly not a random sample, these episodes provide a more grounded opportunity to adjudicate between competing claims about the nature of privately driven legal innovation and the degree of political control within the system. 1. Private Legal Innovation and the Legislative Connection: Pharma- ceutical “Average Wholesale Price” Claims. — An illuminating starting point in exploring the nature of private legal innovation within the qui tam regime is a cluster of suits first filed in the mid 1990s asserting that pharmaceutical companies fraudulently manipulated the “average wholesale price” (“AWP”) of outpatient drugs as a way to boost Medicare and Medicaid reimbursement rates. Such suits arose out of federal laws requiring pharmaceutical companies to disclose the AWP at which they sell their drugs so that government agencies can establish benchmarks the likelihood” that other seven laboratories might also engage in practice); Cooper v. Blue Cross & Blue Shield of Fla., Inc., 19 F.3d 562, 566–67 (11th Cir. 1994) (declining to apply public-disclosure bar as result of reports of industry-wide fraud because of industry’s large and sprawling nature). 161. See supra note 153 and accompanying text (noting imposition of “materiality” requirement via case law and congressional amendments). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1969 for setting prices and rebates under Medicare, Medicaid, and related federal programs. This benchmarking regime creates obvious opportunities for profitable manipulation of the content of those disclosures, as even small shifts in the “spread” between sale and reimbursement prices make drugs relatively more profitable for physicians to prescribe, thus boosting a drug company’s market share and revenue stream. By “marketing the spread” between actual acquisition prices paid by healthcare providers and reimbursement rates, as qui tam relators asserted in a series of lawsuits beginning in 1995, pharmaceutical companies induced healthcare providers and hospitals to submit false claims within the meaning of the FCA.162 AWP suits have generated recoveries totaling at least $3 billion, making them one of the most successful flavors of qui tam litigation.163 Had the qui tam privateers who filed the initial AWP suits merely surfaced information about the various mechanisms drug companies were using to manipulate price spreads in clear violation of governing law, then AWP claims could not be considered legally innovative in any meaningful sense. Nor, had the AWP relators brought to light patently illegal conduct that had escaped regulators’ attention, would the lawsuits raise obvious political or democratic control concerns. In both instances, the AWP lawsuits would instead fit the classic whistleblower model in which private enforcers usefully supplement public enforcement efforts by dislodging privately held information about illegal conduct that public enforcers can obtain only by paying high information costs.164

162. See Krause, Regulating, supra note 9, at 266–67 (explaining theory in AWP cases whereby pharmaceutical companies induce healthcare providers to make false claims). For a listing of the initial AWP suits settled by 2003, see infra note 179. 163. For leading examples, see, e.g., Press Release, U.S. Dep’t of Justice, AstraZeneca Pharmaceuticals LP Pleads Guilty to Healthcare Crime; Company Agrees to Pay $355 Million to Settle Charges (June 20, 2003), http://www.justice.gov/opa/pr/2003/June/03_ civ_371.htm (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Bristol-Myers Squibb to Pay More than $515 Million to Resolve Allegations of Illegal Drug Marketing and Pricing (Sept. 28, 2007), http://www.justice.gov/opa/pr/2007/September /07_civ_782.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Pharmaceutical Manufacturer to Pay $280 Million to Settle False Claims Act Case (Dec. 20, 2010), http://www.justice.gov/opa/pr/2010/December/10-civ-1464.html (on file with the Columbia Law Review) (announcing settlements with Dey Inc., Dey Pharma L.P., and Dey L.P. Inc.); Press Release, U.S. Dep’t of Justice, Pharmaceutical Manufacturers to Pay $421.2 Million to Settle False Claims Act Cases (Dec. 7, 2010), http://www.justice.gov/opa/pr/2010/December/10-civ-1398.html (on file with the Columbia Law Review) (announcing settlements with Roxane Laboratories, Inc., Abbott Laboratories, Inc., and B. Braun Medical Inc.); Press Release, U.S. Dep’t of Justice, TAP Pharmaceutical Products Inc. and Seven Others Charged with Health Care Crimes; Company Agrees to Pay $875 Million to Settle Charges (Oct. 3, 2001), http://www.justice. gov/opa/pr/2001/October/513civ.htm (on file with the Columbia Law Review). 164. See Engstrom, Harnessing, supra note 18, at 1250 (distinguishing “classic whistleblowing insiders with firsthand knowledge of fraud” from “outsider enforcers who develop fraud cases by other means”); see also Yochai Benkler, A Public Accountability 1970 COLUMBIA LAW REVIEW [Vol. 114:1913

Yet two aspects of the AWP suits place them well outside this model and make plain their “interstitial” nature within Part I’s conceptualiza- tion of that term. First, the possibility of AWP manipulation had not gone unnoticed by government regulators at the time of the first qui tam suits asserting AWP claims in 1995.165 To the contrary, a variety of official state- ments and agency reports reaching back more than twenty years before then shows widespread government awareness of persistent gaps between published and actual drug prices.166 Still other evidence likewise suggests pervasive knowledge among regulators and regulated: As early as 1994, a widespread industry joke held that AWP stood for “Ain’t What’s Paid.”167 And in 1997, before many qui tam suits asserting AWP claims were underway, President Clinton took up the issue in his weekly radio address, bluntly announcing that “our Medicare program has been systematically overpaying doctors and clinics for prescription drugs.”168 Second, federal law governing AWP at the time the first qui tam suits were filed was not so much legally ambiguous as it was a gaping legal void.169 Indeed, the Medicare statute and implementing regulations did not provide any substantive restrictions on how drug companies were to calculate AWP, leaving Medicare contractors broad latitude to base their calculations on data derived from voluntary company submissions and collected in privately maintained pharmaceutical pricing compendia.170

Defense for National Security Leakers and Whistleblowers, 8 Harv. L. & Pol’y Rev. 281, 315 (2014) (referring to “classic, internal whistleblowing”). 165. See Krause, Regulating, supra note 9, at 267–68 (noting AWP problem “ha[s] long been known”); Torborg, supra note 118, at 191 (noting AWP problem “was generally well known”). 166. As far back as 1974, Department of Health, Education, and Welfare officials had expressed concern about spreads between published and actual prices. See Krause, Regulating, supra note 9, at 268 (discussing history of discrepancies between published and actual prices); see also Medicare Program, Fee Schedule for Physicians’ Services, 56 Fed. Reg. 25,792, 25,800 (June 5, 1991) (“[T]he Red Book and other wholesale price guides substantially overstate the true cost of drugs.”); Office of the Inspector Gen., Dep’t of Health & Human Servs., A-06-97-00011, Medicaid Pharmacy—Actual Acquisition Cost of Generic Prescription Drug Products 5 (Aug. 1997) (finding generic drugs sold at average discount of 42.5 percent off reported AWP); Torborg, supra note 118, at 190 n.38 (discussing problem of pharmaceutical discounting). 167. See, e.g., Mick Kolassa, Guidance for Clinicians in Discerning and Comparing the Price of Pharmaceutical Agents, 9 J. of Pain & Symptom Mgmt. 235, 237 (1994) (noting joke). 168. See The President’s Radio Address at the White House, 33 Weekly Comp. Pres. Doc. 2033–34 (Dec. 13, 1997). 169. See Torborg, supra note 118, at 192 (“[T]here was no statute or regulation defining or establishing duties upon manufacturers in connection with AWP....”). 170. See John K. Iglehart, Medicare and Drug Pricing, 348 New Eng. J. Med. 1590, 1591 (2003) (describing industry’s use of pricing compendia). At the time of the first AWP suits, the Medicare statute and implementing regulations reimbursed physicians based on actual charges or ninety-five percent of AWP, but did not define the term, leaving Medicare contractors to base their calculations on data collected in pharmaceutical pricing publications reporting voluntary company submissions. The most popular of these 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1971

Moreover, this legal void had persisted despite multiple efforts during the George H.W. Bush and Clinton Administrations to legislate a new approach.171 So difficult was it to form a legislative consensus that Congress did not just fail to enact a solution; Congress also stymied proposed administrative fixes, imposing a moratorium on efforts by the Health Care Financing Administration (HCFA) to develop an alternative source of AWP data pending a comprehensive study of the issue.172 As a result, it was not until 2003—some eight years after the first AWP cases were filed and two years after the DOJ, having initially declined to join a case against TAP Pharmaceuticals, intervened and pushed it to a landmark settlement173—that the Centers for Medicare & Medicaid Services (CMS, previously HCFA) first published a notice of proposed rulemaking announcing revisions to the drug reimbursement methodology and Congress began serious debate about an AWP fix.174 Soon after, Congress passed the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA), overhauling the drug-pricing system by replacing AWP with an “average sales price” approach.175 pricing compendia was the Drug Topics Red Book. See Michael L. Koon, AWP Through the Looking Glass: Industry Litigation by Definition, For the Defense (Def. Research Inst., Chi., Ill.), Oct. 2005, at 45, 47, available at http://www.shb.com/practiceareas/Pharma/ Pubs/IndustrywideLitigation_2005.pdf (on file with the Columbia Law Review) (describing origin of AWP in privately maintained Red Book). 171. See Krause, Regulating, supra note 9, at 268 (noting both Administrations failed to substantially decrease AWP). Though various possible overhauls were debated beginning just after Ven-A-Care’s maiden lawsuit in 1995, Congress could only reach consensus on a reduction in reimbursement rates to ninety-five percent of AWP as a way to cut into the surplus, which critics said the industry was achieving via AWP manipulation. See Balanced Budget Act of 1997, Pub. L. No. 105-33, § 4556(a), 111 Stat. 251, 462–63 (amending 42 U.S.C. § 13958(o)(1) (1994)) (enacting ninety-five-percent reimbursement rate); see also Iglehart, supra note 170, at 1591 (describing lack of consensus about how to revise pricing methodology). 172. Specifically, in the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000, Congress provided some authority for HHS to revise the reimbursement methodology, but not until the GAO had completed a report on the subject. See, e.g., Medicare Drug Reimbursements: A Broken System for Patients and Taxpayers, Hearing Before the Subcomm. on Health and Subcomm. on Oversight & Investigations of the H. Comm. on Energy & Commerce, 107th Cong. 55 (2001) (describing congressional moratorium). 173. See United States v. TAP Holdings, No. 1:98-CV-10547 (D. Mass. filed Sept. 28, 2001) (showing, on docket sheet, United States intervening in late 2001); Krause, Following, supra note 114, at 361 (describing DOJ entry into AWP cases); Press Release, U.S. Dep’t of Justice, TAP Pharmaceutical Products, Inc. and Seven Others Charged with Health Care Crimes; Company Agrees to Pay $875 Million to Settle Charges (Oct. 3, 2001), http://www.usdoj.gov/opa/pr/2001/October/513civ.htm (on file with the Columbia Law Review) (setting forth terms of $875 million settlement with TAP Pharmaceuticals). 174. See Medicare Program: Payment Reform for Part B Drugs, 68 Fed. Reg. 50,428, 50,431–35 (proposed Aug. 20, 2003) (proposing four different options for revising Medicare Part B payment methodology). 175. See Cong. Research Serv., RL31199, Medicare: Payments to Physicians 25 (2008) 1972 COLUMBIA LAW REVIEW [Vol. 114:1913

Why the prolonged congressional inertia on reimbursement methodology, and why no separate agency or prosecutorial efforts to use available enforcement tools, including government-filed FCA actions, to move against AWP manipulation until well after the first qui tam suits were filed? The most likely explanation is that the reimbursement- methodology issue was especially politically fraught. A substantial proportion of the Medicare-eligible drugs subject to AWP benchmarking were cancer treatments, and a number of commentators have explained that Congress and HHS dared not touch the reimbursement issue for fear of incurring the “wrath of the oncology lobby . . . [and] other powerful physicians’ groups,” which had long argued that the extra cost of admin- istering cancer drugs justified wide spreads between actual and reported prices.176 On this view, qui tam enforcement efforts asserting AWP claims exploited a statutory and regulatory vacuum created by the third-rail political nature of cancer treatment. The AWP lawsuits thus raise troubling questions about the tendency of qui tam enforcers to exploit legislative inertia and, by extension, the overall degree of political control within the system. As Part I noted, interstitial private enforcement efforts may not draw legislative override.177 The AWP example, with congressional action on drug reimbursement coming some eight years after the first qui tam suits were filed, would thus seem to reflect Part I’s logic in which private enforcement efforts operate within a “discretion” or “gridlock” space wherein political control is less dependable. Less clear is whether, as Part I also hypothesized, qui tam suits asserting AWP claims reshaped the political context in ways that pushed the evolution of drug-reimbursement regulation off the path it would have traveled in the absence of a qui tam mechanism. Part of the challenge here is methodological: As noted previously, counterfactual questions of this sort rarely admit of definitive answers. Even so, it is noteworthy that the paper trail from Congress’s legislative deliberations leading up to enactment of the MMA in 2003 is devoid of any evidence that could be

(discussing average sales price methodology adopted by MMA). Among other things, the Medicare Modernization Act of 2003 (MMA) reduced reimbursement to eighty-five percent of AWP for 2004 and put into place a new payment mechanism for 2005 under which payment for single-source drugs would be based on the lesser of the statutorily defined “average sales price” or the “wholesale acquisition cost.” See Medicare Modernization Act, Pub. L. No. 108-173, § 303(c), 117 Stat. 2066, 2240–43 (2003) (codified as amended at 42 U.S.C. § 1395w-3A(b)(4) (2012)). 176. Krause, Regulating, supra note 9, at 269; see also Iglehart, supra note 170, at 1590 (describing oncology lobby’s opposition to reform). For a representative example of oncologist complaints during congressional deliberations on the MMA, see Medicare Payments for Currently Covered Prescription Drugs: Hearing Before the Subcomm. on Health of the H. Comm. on Ways & Means, 107th Cong. 3 (2002). 177. See supra notes 77–81 and accompanying text (summarizing mechanisms that can stymie legislative override of private legal innovations deviating from legislative purposes). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1973 construed as supporting Part I’s hypothesized dynamic in which early litigation losers lobby to ensure their competitors suffer the same fate.178 Indeed, a thorough review of the more than twenty days of floor debate and committee hearings reveals no evidence of cleavages between industry actors that had paid out large settlements—including, by 2003, TAP Pharmaceuticals, AstraZeneca, and Bayer—and those that had not.179 In some ways, this lack of evidence of a feedback effect is surprising. Like Part I’s hypothesized Wal-Mart-like defendant, which had made irreversible changes to its managerial structure in response to a litigated judgment against it, the three pharmaceutical companies that had paid out AWP settlements by 2003 all faced strong incentives to ensure that their competitors faced the same liability they had faced. In particular, the settlements, in addition to requiring the companies to pay substantial damages, included Corporate Integrity Agreements (CIAs) compelling use of average sales prices, net of discounts or rebates, rather than AWP going forward.180 This mandate left these companies with few good lobbying options as the rulemaking and legislative efforts got underway in 2003 other than to advocate that their competitors be subject to a similarly unforgiving approach, and ideally one with retroactive effect. After all, an administrative or congressional fix setting into place a more forgiving reimbursement methodology might abrogate the settlement’s provisions mandating a stricter reimbursement methodology, thus maximizing the settling companies’ revenues going forward. But such a fix would also maximize their competitors’ revenues while undercutting the ongoing litigation against those same competitors for the pre-2003

178. For a selection of the congressional deliberations leading up to the MMA’s passage, see Waste, Fraud, and Abuse: Hearing Before the H. Comm. on Ways & Means, 108th Cong. (2003); H.R. 1, Medicare Prescription Drug and Modernization Act of 2003: Hearing Before the H. Comm. on Rules, 108th Cong. (2003); Medicare Reimbursements for Physicians and Hospitals: Hearing Before the Subcomm. on Labor, Health, & Human Servs., Educ., & Related Agencies of the S. Comm. on Appropriations, 108th Cong. (2003); 149 Cong. Rec. 16,281, 16,440, 30,959, 31,121, 31,780, 31,817–18 (2003) (providing sample of congressional deliberations and discussions on passage of MMA). 179. The three AWP cases settled by 2003 include (i) the $875 million settlement with TAP Pharmaceuticals noted previously, see supra note 173 and accompanying text; (ii) the $280 million settlement with AstraZeneca, see Press Release, U.S. Dep’t of Justice, AstraZeneca Pharmaceuticals LP Pleads Guilty to Healthcare Crime; Company Agrees to Pay $355 Million to Settle Charges (June 20, 2003), http://www.justice.gov/archive/opa/ pr/2003/June/03_civ_371.htm (on file with the Columbia Law Review); and (iii) the comparatively smaller $14 million settlement with Bayer, see Press Release, U.S. Dep’t of Justice, Bayer to Pay $14 Million to Settle Claims for Causing Providers to Submit Fraudulent Claims to 45 State Medicaid Programs (Jan. 23, 2001), http://www.justice.gov/ opa/pr/2001/January/039civ.htm (on file with the Columbia Law Review). 180. See, e.g., Corporate Integrity Agreement Between the Office of Inspector General of the Department of Health and Human Services and TAP Pharmaceutical Products, Inc. 8–11 (Sept. 28, 2001), available at http://oig.hhs.gov/fraud/cia/ agreements/tap_pharmaceutical_products_92801.pdf (on file with the Columbia Law Review) (compelling TAP Pharmaceuticals to report average sales prices and enumerating other requirements). 1974 COLUMBIA LAW REVIEW [Vol. 114:1913 period by making available the powerful defense that Congress had since sanctioned the prior methodology and disclosures. In other ways, however, the evidentiary void should not surprise at all. It is well established that much lobbying on regulatory matters occurs behind the scenes, outside of public legislative-committee hearings or the docketed notice-and-comment process that agencies conduct under the APA.181 Early-settling companies seeking to prop up the ongoing AWP suits against their competitors might be especially likely to make their case behind closed doors to avoid having their efforts branded as anticompetitive or unprincipled. This might explain why the deliberations leading up to the MMA’s passage featured few industry voices at all—and were instead dominated by nonprofit organizations, including the Urban Institute, Taxpayers Against Fraud, the United Council on Welfare Fraud, and the American Association of Retired Persons, advancing patient- and taxpayer-friendly positions.182 It is also important to recall that the initial AWP suits in the mid- 1990s and the further wave of filings in 2001 and 2002 had, by the time CMS and Congress began to consider revisions to the pricing regime in 2003, produced only three settlements.183 Indeed, the bulk of the billions in settlements came well after, during the period from 2009 to 2011.184

181. There is no proscription on ex parte communication in the legislative process, and the APA has been interpreted to place only weak constraints on ex parte lobbying in the informal rulemaking process. See, e.g., Sierra Club v. Costle, 657 F.2d 298, 400–01 (D.C. Cir. 1981) (asserting that barring ex parte contacts in informal rulemaking would be of “questionable utility” and requiring agency to docket contacts only where it intends to rely on them to justify rule). 182. As just one example, an October 2002 hearing before the Subcommittee on Health of the powerful House Committee on Ways and Means featured testimony about AWP by a wide variety of provider organizations (e.g., American Society of Clinical Oncology, American Association for Homecare, Association of Freestanding Radiation Oncology Centers, and National Alliance for Infusion Therapy) and patient-focused organizations (e.g., Project HOPE and Medical Rights Center) but no testimony at all by pharmaceutical companies. See Medicare Payments for Currently Covered Prescription Drugs, Hearing Before the Subcomm. on Health of the H. Comm. on Ways & Means, 107th Cong. iii (2002) (listing organizations offering testimony at hearing). 183. See supra note 179 and accompanying text (describing TAP Pharmaceuticals, AstraZeneca, and Bayer settlements). 184. Some of the largest settlements, most of them brought about by Ven-A-Care, did not occur until at least 2007, beginning with a $515 million settlement with Bristol-Meyers Squibb, see Press Release, U.S. Dep’t of Justice, Bristol-Myers Squibb to Pay More than $515 Million to Resolve Allegations of Illegal Drug Marketing and Pricing (Sept. 28, 2007), http://www.justice.gov/opa/pr/2007/September/07_civ_782.html (on file with the Columbia Law Review), and continuing on to nearly $2 billion in settlements. See Thomas Catan, Drug Makers Agree to $421 Million Settlement, Wall St. J. (Dec. 8, 2010, 12:01 AM) (discussing Ven-A-Care’s involvement in $421 million settlement against three drug makers in 2007); Thomas Catan, Cha-Ching! Payouts Continue for Four Florida Whistleblowers, Wall St. J.: L. Blog (Dec. 8, 2010, 4:10 PM), http://blogs.wsj.com/law/ 2010/12/08/cha-ching-payouts-continue-for-four-florida-whistleblowers/ (on file with the Columbia Law Review) (noting Ven-A-Care has “brought suit after suit under the qui tam provisions” and has “helped recover more than $2 billion for the U.S. government”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1975

Given that so few settlements had been concluded at the time CMS and Congress entered the fray in 2003, it may be that lobbying efforts to ensure that the rest of the pharmaceutical industry would face similar liability were simply not intense enough to register in the legislative history. To that extent, the AWP lawsuits present a conservative case for finding clear evidence of a positive-feedback or increasing-returns mechanism of the sort Part I hypothesized. Yet even if the AWP episode offers few definitive answers on the plausibility of Part I’s feedback dynamic, it still permits some initial generalizations about the degree of public control within private enforcement regimes. For instance, working through the AWP example suggests that the strength of the feedback loop created by private litigation is likely to turn on whether litigation efforts take omnibus, pan-industry forms or instead play out in a sequence of smaller-scale or even single- defendant suits. In pan-industry suits, judgments and settlements are likely to come relatively close in time, leaving less room for industry actors who suffer early defeats to lobby to ensure their competitors are exposed to the same liability in ways that yield interpretive pathways that differ from a world without private enforcement. Thus, contrary to many of qui tam’s critics185—and also broader critics of “regulation by litigation”186—private enforcement efforts targeting much or all of an industry in one fell swoop may, perhaps counter-intuitively, present a lower risk of statutory drift than multiple, sequential litigation efforts.187 A second generalizable insight is that questions about the degree of public control within private enforcement regimes do not just implicate complex counterfactuals about how litigation losses impact subsequent legislative efforts. They also implicate separation-of-powers dynamics. Recall that the first large AWP case reached settlement only after the DOJ, having initially declined to become involved, belatedly intervened and negotiated a settlement. This is important because the DOJ’s presence in the earliest AWP settlements permits an interpretation of the episode as an executive or prosecutorial end-run around the legislative and administrative stalemate rather than an instance of privately driven, publicly unaccountable legal innovation.188 On this view, private enforce-

185. See supra notes 9, 118–121 and accompanying text (detailing criticisms of qui tam litigation as politically unaccountable). 186. See supra note 70 and accompanying text (discussing definition and criticisms of “regulation by litigation”). 187. See supra notes 73–83 and accompanying text (discussing risk of statutory drift posed by private enforcement). 188. See, e.g., Joan H. Krause, A Conceptual Model of Health Care Fraud Enforcement, 12 J.L. & Pol’y 55, 110–37 (2003) (describing use of litigation to fill regulatory gaps and problems inherent in such approach); Krause, Following, supra note 114, at 360 (“[T]he TAP settlement was designed to accomplish what the legislative and regulatory processes had . . . failed to do: generate a more accurate source of drug pricing information. The DOJ and OIG . . . had used the litigation process to begin to close a persistent and well-known loophole in Medicare law and regulations.”). 1976 COLUMBIA LAW REVIEW [Vol. 114:1913 ment efforts like the AWP cases provide political cover for an executive or prosecutorial power grab or otherwise serve as a pawn in a broader separation-of-powers or interagency struggle, but they are not the principal threat to public control.189 Such an interpretation also highlights the fact that the FCA regime, like much of the American regulatory state, is neither a purely public nor a purely private regime but rather a complex public–private hybrid.190 This is why an empirical understanding of the DOJ’s exercise of its statutory gatekeeper powers, as Part IV.C provides, is necessary to fully evaluate the political-control concerns voiced by qui tam’s critics. 2. Private Legal Innovation and the Bureaucratic Connection: Oil- and Gas-Royalty Claims. — If the AWP cases offer an opportunity to explore political- and democratic-control concerns where private qui tam enforcers exploit legislative inertia, then other examples from outside the healthcare context further illuminate the complicated bureaucratic connection within the FCA regime. Consider in this regard a second significant flavor of qui tam litigation and a perennial target of FCA critics: qui tam actions targeting much or all of the oil and gas industry for underpayment of royalties in connection with extraction operations on federal and Indian lands.191 As with the AWP suits, royalty-valuation claims constitute a nontrivial fraction of total qui tam enforcement activity, generating some 250 filings and as much as $1.1 billion in recoveries since 1986.192 While most qui tam cases alleging royalty underpayment level allegations at individual defendants, several of the more successful cases have been omnibus complaints, asserting that the entire oil and gas industry has fraudulently utilized a common set of accounting or measurement gimmicks to shirk its payment obligations. One such example is United

189. See, e.g., Ashutosh Bhagwat, Modes of Regulatory Enforcement and the Problem of Administrative Discretion, 50 Hastings L.J. 1275, 1304–05 (1999) (noting general concern about lawmaking through enforcement); Marc J. Scheineson & Shannon Thyme Klinger, Lessons from Expanded Government Enforcement Efforts Against Drug Companies, 60 Food & Drug L.J. 1, 7 (2005) (criticizing government’s increased use of fraud statutes “to force settlements by drug and device manufacturers for conduct that was, in large part, viewed by FDA and other agencies as acceptable industry practice until the DOJ and OIG began to redefine the regulatory landscape”). 190. See Engstrom, Gatekeepers, supra note 16, at 704 (describing “current qui tam regime under the FCA” as “fully ‘hybrid’ public-private enforcement model”); Krause, Regulating, supra note 9, at 273 (noting FCA is different from other contexts that have drawn “regulation by litigation” criticisms because of “dual private-public enforcement mechanism embodied in the FCA”). 191. Federal lands subject to royalty include onshore federal lands throughout the fifty states (though the lower forty-eight and Alaska and Hawaii are subject to separate regimes) and offshore lands, including the Outer Continental Shelf (OCS) of the Gulf of Mexico and Pacific. For an example of criticism of oil and gas qui tam cases, see Salcido, Government Declares War, supra note 118, at 4 (discussing relator interest in maximizing personal gain and noting provision of United States funds for “useless information”). 192. This figure is calculated from the data described supra Part II.C. 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1977

States ex rel. Johnson v. Exxon Co. USA, where the relator—a former oil- industry executive193—achieved settlements totaling nearly half a billion dollars across sixteen different oil and gas companies based on allegations that virtually the entire industry had deliberately undervalued oil production by, among other things, using prices derived from non-arms- length sales to affiliated entities as valuation benchmarks in violation of governing regulations.194 A second example is United States ex rel. Wright v. Amerada Hess, in which the relator’s initial complaint alleged industry- wide use of various artifices to evade royalty obligations but then largely focused, after substantial litigation, on a single allegation: that the defendants had failed to pay royalties on raw natural gas diverted from processing to fuel plant machinery that “boosted” the pressure of processed gas for injection into interstate pipelines for delivery to market.195 Like the Johnson litigation, the Wright litigation ultimately yielded substantial settlements with a large proportion of the industry, with total payouts exceeding $300 million across more than a dozen companies.196

193. See J. Benjamin Johnson, Jr., Summit Res. Mgmt., Inc., Professional Resume’ [sic] of J. Benjamin Johnson, Jr., http://summitrm.com/JBJ%20Resume.htm (on file with the Columbia Law Review) (last visited Oct. 7, 2014) (listing work experience of J. Benjamin Johnson, Jr., including over a decade of employment with ARCO Oil and Gas Company). 194. See Original Complaint and Jury Demand at 24, United States ex rel. Johnson v. Exxon Co. USA, No. 9:96-cv-00066-TH (E.D. Tex. Feb. 16, 1996), 1996 WL 34303209 (alleging defendants used “affiliated companies who purchase the oil for value less than the producing company could have sold it in a true arms-length transaction”). The basic idea here is that an oil and gas company that wishes to minimize royalty payments can engage in underpriced transfers of the oil and gas to a related entity under its corporate control prior to selling it onto the open market and then use the lower price from the affiliate transaction, not the higher price commanded on the open market, as the basis for its payment of royalties to the United States. In particular, the Johnson complaint alleges that oil and gas companies engaged in such transactions without adjusting the reported prices using alternative valuation benchmarks required by governing regulations. See 30 C.F.R. § 206.102 (1995) (discussing oil- and gas-royalty valuation, including alternative valuation benchmarks). The “nearly half a billion” figure is derived from the data and methods described supra Part II.C. See also U.S. Dep’t of Justice, The Accomplishments of the U.S. Department of Justice, 2001–2009, at 11, available at http://www.justice.gov/sites /default/files/opa/legacy/2010/03/08/doj-accomplishments.pdf (on file with the Columbia Law Review) (last visited Oct. 7, 2014) (noting settlements totaling $440 million with sixteen oil companies). 195. Interestingly, the initial complaint in Wright did not specifically mention the “boosting” claim. Complaint at 22–28, United States ex rel. Wright v. Chevron, U.S.A., No. 5:96-cv-00243 (E.D. Tex. Aug. 2, 1996). The gist of the claim is that fully processed gas must be brought up to high pressure in order to inject it into the nation’s interstate pipelines for delivery to market. But doing so is expensive, as the compressors necessary to “boost” gas up to pipeline pressure consume substantial energy themselves. In the Wright litigation, the relator claimed that defendants were diverting raw gas from processing to fuel those compressors, thus powering their operations with royalty-free gas. See, e.g., Mobil Defendants’ Motion for Summary Judgment on Relator’s Claims at 13, United States ex rel. Wright v. Chevron, U.S.A., No. 5:03-cv-00264-MHS-CMC (E.D. Tex. May 1, 2008) (describing “boosting” claim). 196. See, e.g., Press Release, U.S. Dep’t of Justice, Anadarko and Kerr-McGee to Pay More than $17 Million to Resolve Allegations of Royalty Underpayments from Federal and 1978 COLUMBIA LAW REVIEW [Vol. 114:1913

Royalty-valuation suits like Johnson and Wright resemble the AWP cases and raise many of the same concerns about “regulation by litigation” in which private enforcers exploit regulatory ambiguities in industry-wide actions that achieve regulatory ends normally reserved for ex ante legislative action or agency rulemaking.197 Most obviously, neither suit fits the classic whistleblower model in which company insiders surface information about wrongdoing that is prohibitively costly for public regulators to discover or dislodge. Rather, the underpayment claims alleged in each case were plainly interstitial within Part I’s meaning. For example, in the Wright litigation, the “boosting” claims at the core of many of the settlements targeted an ambiguity in federal royalty regulations as to whether compression at the tailgate of the plant was within the confines of the plant, and thus subject to royalty as part of the “production” process, or outside the plant, and thus exempt from royalty as part of the “marketing” process.198 Furthermore, the practices challenged in both cases were arguably well known to government regulators—with defendants adducing evidence demonstrating at least some degree of prior regulator notice.199 In both ways, AWP and royalty- valuation suits are plainly of a piece.

Indian Lands (June 20, 2011), http://www.justice.gov/usao/txe/News/2011/11-795.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, BP Amoco to Pay U.S. $20.5 Million to Resolve Allegations of Royalty Underpayments from Federal and Indian Lands (Sept. 16, 2011), available at http://www.justice.gov/opa/pr/2011/ September/11-civ-1201.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Chevron to Pay U.S. More than $45 Million to Resolve Allegations of False Claims for Royalties Underpayment (Dec. 23, 2009), available at http://www.justice.gov/ opa/pr/2009/December/09-civ-1379.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Devon Energy to Pay U.S. $3.5 Million to Resolve Allegations of Royalty Underpayments from Federal and Indian Lands (Mar. 12, 2012), available at http://www.justice.gov/opa/pr/2012/March/12-civ-309.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Dominion and Marathon Oil to Pay More than $6.9 Million to Resolve Allegations of Royalty Underpayments from American Indian and Federal Lands (Aug. 20, 2010), available at http://www.justice.gov/opa/pr/2010/ August/10-civ-942.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Occidental Oil Companies to Pay $2.05 Million to Resolve Allegations of Royalty Underpayments from American Indian and Federal Lands (Mar. 22, 2011), available at http://www.justice.gov/opa/pr/2011/March/11-civ-363.html (on file with the Columbia Law Review); Press Release, U.S. Dep’t of Justice, Total Companies to Pay U.S. $15 Million to Resolve Allegations of Royalty Underpayments from Federal and Indian Lands (Feb. 2, 2012), available at http://www.justice.gov/opa/pr/2012/February/12-civ-240.html (on file with the Columbia Law Review). 197. See supra notes 9, 118–121 and accompanying text (discussing various criticisms of qui tam litigation, including relators’ tendency to “exploit regulatory gray areas rather than surfacing plain-vanilla, ‘raw’ frauds”). 198. See Mobil Defendants’ Motion for Summary Judgment on Relator’s Claims at 13, Wright, No. 5:03-cv-00264-MHS-CMC (describing “boosting” claim); 30 C.F.R. § 206.102 (detailing royalty valuation for oil sold “pursuant to an arms-length contract”). 199. See, e.g., United States ex rel. Johnson v. Shell Oil Co., et al., 33 F. Supp. 2d 528, 538 (E.D. Tex. 1999) (noting defendants’ argument alleging schemes were common industry practices well known to government regulators and fully contemplated by royalty 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1979

Yet the Johnson and Wright cases also resemble the AWP cases in a final sense that once more highlights the difficulty of drawing firm conclusions about the ultimate degree of political or democratic control. In particular, neither case offers concrete evidence of Part I’s feedback dynamic in which an initial set of judgments against key industry actors reshapes the interest-group environment and thus drives legal development in directions it would not go in the absence of private litigation.200 Indeed, the legislative and administrative record reveals no evidence that Congress or the Department of the Interior ever considered revising the regulations regarding arms-length sales or “boosting.”201 The result is an interpretive impasse: Perhaps congressional and administrative silence on the boosting issue provides evidence against a feedback dynamic. Or perhaps it shows that the companies that entered into early settlements successfully forestalled pro-industry statutory or regulatory revisions, so that their competitors would be subject to the same liability. It is still possible, however, to derive some midlevel generalizations from the royalty-valuation suits like Johnson and Wright, particularly when compared to the AWP cases. Indeed, note a key difference between the two: While Congress affirmatively forbade administrative efforts to fix the AWP-reimbursement methodology, Congress had long delegated wholesale authority over royalty collection to the Secretary of the Interior and, by further delegation, the Bureau of Land Management (BLM) and Minerals Management Service (MMS) for onshore and offshore extraction, respectively.202 To that extent, royalty-valuation suits have arguably regulations); Mobil Defendants’ Motion for Summary Judgment on Relator’s Claims at 5, Wright, No. 5:03-cv-00264-MHS-CMC (citing testimony by MMS’s former head of enforcement that MMS audits conducted four years before Wright case was filed focused on challenges to recompression—i.e., “boosting”—of natural gas); Shell Defendants’ Overview Memorandum in Support of Motions for Partial Summary Judgment and Exclusion of Vogel Expert Report at 1, Johnson, No. 9:96-cv-00066-TH (E.D. Tex. Aug. 28, 2000) (noting Shell’s principal defense was that it achieved express agency direction via value determination letter with respect to significant portion of oil production at issue). 200. See supra note 98 and accompanying text (discussing feedback effect of early litigation outcomes on ultimate outcomes). 201. The extensive congressional hearings that touched on oil- and gas-royalty issues following revelations about inappropriate conduct at the Minerals Management Service, see infra note 204, are silent on the “boosting” issue. See, e.g., Getting Royalties Right: Recent Recommendations for Improving the Federal Oil and Gas Royalty System: Hearing Before the Subcomm. on Energy & Mineral Res. of the H. Comm. on Natural Res., 110th Cong. (2008); Oversight of the Oil, Gas, and Mineral Revenue Programs Managed by the Department of the Interior: Hearing Before the Subcomm. on Dep’t of the Interior, Env’t, & Related Agencies of the S. Comm. on Appropriations, 110th Cong. (2008); Natural Gas Royalties: The Facts, The Remedies: Hearing Before the Subcomm. on Energy & Res. of the H. Comm. on Gov’t Reform, 109th Cong. (2006); Deep Water Royalty Relief: Mismanagement and Cover-Ups: Hearing Before the Subcomm. on Energy & Res. of the H. Comm. on Gov’t Reform, 109th Cong. (2006). 202. Oil and gas royalties for onshore areas are governed by the Federal Oil and Gas Royalty Management Act of 1982, Pub. L. No. 97-451, § 101(a), 96 Stat. 2447, 2449 (codified as amended at 30 U.S.C. § 1711(a) (2012)), which delegated authority to BLM 1980 COLUMBIA LAW REVIEW [Vol. 114:1913 tended to exploit legal ambiguities left by administrative, not legislative, inertia. This is important, for oil- and gas-royalty-valuation cases have gone forward amid longstanding concern that one of the principal agencies charged with royalty collections, the MMS within the Department of the Interior, was dysfunctional—or, worse, was “captured” by industry.203 Indeed, as the Wright case was in active litigation, the Office of the Inspector General (OIG) of the Department of the Interior released several bombshell reports reporting serious conflict-of-interest issues, including a pattern of gifts and improper sexual relationships between MMS officials and oil- and gas-industry personnel.204 The more recent Deepwater Horizon explosion in the Gulf of Mexico only deepened such concerns and, soon after, led to the agency’s dissolution.205 The royalty-valuation cases thus offer an extreme example of a more general point: Private qui tam enforcement efforts targeting administrative, as opposed to legislative, inertia might be thought less concerning from a political- or democratic-control perspective—or, at the least, raise even more complicated separation-of-powers considerations than those arising in the AWP cases. In particular, one could argue that the relators in the Johnson and Wright cases played a salutary agency-forcing or anti- to establish “comprehensive inspection, collection and fiscal and production accounting and auditing system.” Offshore gas royalties are governed by the Outer Continental Shelf Lands Act, Pub. L. No. 212, 67 Stat. 462 (1953) (codified as amended at 43 U.S.C. § 1334(a) (2012)). 203. See supra note 31 and accompanying text (reviewing capture concept). 204. See Memorandum on OIG Investigations of MMS Employees from Earl E. Devaney, U.S. Inspector Gen., to Dirk Kempthorne, U.S. Sec’y of the Interior (Sept. 9, 2008), available at http://www.doi.gov/oig/reports/upload/SmithGregory-2008-09-10.pdf (on file with the Columbia Law Review) (summarizing results of three separate investigations focused on MMS activities during 2002 through 2006). MMS has also been the subject of withering reports by the Government Accountability Office and Congress itself. See U.S. Gov’t Accountability Office, GAO-09-744, Royalty-in-Kind Program: MMS Does Not Provide Reasonable Assurance It Receives Its Share of Gas, Resulting in Millions in Forgone Revenue 1–2 (2009), available at http://www.gao.gov/new.items/d09744.pdf (on file with the Columbia Law Review) (discussing failure to collect revenues owed to federal government); Staff of H. Comm. on Oversight and Gov’t Reform, 111th Cong., Teapot Dome Revisited: Dereliction of Fiduciary Duty at the Interior Department 1–2 (2009), available at http://oversight.house.gov/wp-content/uploads/2012/02/2009-10-07 mmsreport.pdf (on file with the Columbia Law Review) (detailing congressional investigation into MMS abuse and mismanagement). See generally Christopher Carrigan, Captured by Disaster? Reinterpreting Regulatory Behavior in the Shadow of the Gulf Oil Spill, in Preventing Regulatory Capture, supra note 31, at 239, 241–46 (reviewing criticisms leading to MMS’s dissolution). 205. Reorganization of Title 30: Bureaus of Safety and Environmental Enforcement and Ocean Energy Management, 76 Fed. Reg. 64,432 (Oct. 18, 2011) (codified at 30 C.F.R. chs. II, V) (announcing restructuring of MMS into three new bureaus—Bureau of Ocean Energy Management (BOEM), Bureau of Safety and Environmental Enforcement (BSEE), and Office of Natural Resources Revenue (ONRR)—and renaming former MMS to Bureau of Ocean Energy Management, Regulation and Enforcement (BOEMRE)). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1981 capture role in countering MMS lassitude or corruption—and did so in a way that was entirely consistent with Congress’s clearly expressed prefer- ence in the 1986 amendments reviving the FCA that qui tam suits perform precisely this role.206 In the end, the threat that the royalty-valuation cases pose to public control of legal evolution appears to be even less clear-cut than it is in the AWP context. 3. Private Legal Innovation and Regulatory Colonization: Healthcare- Kickback and Self-Referral Claims.— Qui tam enforcement actions asserting healthcare kickback and self-referral claims provide a final illuminating example. These claims key off of statutory and regulatory mandates prohibiting “kickbacks” (i.e., payments offered by healthcare providers to induce referrals for health services or goods paid for by the government) and “self-referrals” (i.e., patient referrals to an entity, such as a clinical laboratory, for health services paid for by the government where the referring physician has an investment or other “financial relationship” with the entity).207 These prohibitions are necessary, policymakers argue, to maintain competitive healthcare markets and reduce unneces- sary care by ensuring that medical judgment, not profit, drives care decisions among providers with joint financial stakes.208 As with AWP

206. During legislative debates leading up to the FCA’s 1986 revival, Congress repeatedly noted that qui tam relators might serve an agency-forcing and anticapture role and even designed a two-tiered system of payouts—in which relators earn a higher bounty when they proceed without the DOJ—in order to incentivize relators to go it alone where a politicized bureaucracy refused to enforce. See Engstrom, Public Regulation, supra note 85, at 1711 n.68 (collecting congressional testimony on 1986 FCA amendments). 207. The “kickback” and “self-referral” concepts are straightforward but the law is complicated, spanning three separate statutes and including criminal and civil prohibitions, with remedies under the latter including civil monetary penalties and exclusion from program participation. Kickback provisions are found in what is commonly referred to as the Anti-Kickback Statute (AKS). The AKS’s criminal provisions were enacted as part of the Social Security Amendments of 1972, Pub. L. No. 92-603, § 242, 86 Stat. 1329, 1419–20 (codified as amended at 42 U.S.C. § 1320a-7b (2012)). The AKS’s civil provisions, sometimes referred to as the Civil Monetary Penalties Law, came later as part of the Medicaid Program and Patient Protection Act of 1987, Pub. L. No. 100-93, § 3, 101 Stat. 680, 686 (codified as amended at 42 U.S.C. § 1320a-7a (2012)). Administrative enforcement of the statute is handled by the Office of the Inspector General of the Department of Health and Human Services, while the DOJ and the U.S. Attorneys’ Offices oversee criminal and civil litigation. The second and third statutes govern self-referrals and are part of the Ethics in Patient Referrals Act, commonly referred to as the Stark I and Stark II laws, named for their congressional sponsor, Fortney “Pete” Stark. Stark I took effect on January 1, 1992, and applied to the referral of Medicare patients for clinical laboratory services. Pub. L. No. 101-239, § 6204, 103 Stat. 2106, 2236–43 (1989) (codified at 42 U.S.C. § 1395nn (2012)). Stark II came via the Omnibus Budget Reconciliation Act of 1993 and expanded Stark I’s self-referral prohibition to Medicaid and ten additional “designated health services,” effective as of December 31, 1994. Pub. L. No. 103-66, §§ 13,562, 13,624, 107 Stat. 312, 596–605, 636 (1993) (codified as amended at 42 U.S.C. § 1395nn (2012)). For a succinct overview of the principal differences between the kickback and self-referral laws, see Matthew, Tainted, supra note 124, at 554. 208. To provide a recent example of conduct implicating antikickback laws, a new wave of qui tam suits targeted insurer efforts to influence physician referrals using, among 1982 COLUMBIA LAW REVIEW [Vol. 114:1913 claims, kickback and self-referral claims have generated more than a billion, and perhaps as much as several billion, dollars in FCA recoveries since 1986.209 As will become clear, qui tam cases asserting kickback and self- referral claims underscore many of the observations made in connection with the AWP and royalty-valuation cases. Among these are the tendency of qui tam enforcement efforts in the AWP and royalty-valuation contexts to exploit regulatory ambiguities in targeting conduct that is well known to the government, and also the complicating presence of separation-of- powers dynamics in any effort to draw clean inferences about the degree of political control over litigation. But considering qui tam kickback suits alongside the AWP and royalty-valuation examples adds unique analytic value in at least two ways. First, the initial kickback suits were not just substantively innovative, but also jurisdictionally innovative. As an initial matter, the first qui tam kickback suits were a species of what some courts have since dubbed “certification” claims: Relators alleged that reimbursement requests made by healthcare providers in violation of kickback laws were rendered “false” within the FCA’s meaning, despite the fact that the providers had other things, “pull-through” schemes in which health insurers, because they could not lawfully prevent certain insureds from seeking “out-of-network” care, instead promised laboratories they would induce physicians via rewards and penalties to steer services to those laboratories in return for discounts. See Charles C. Dunham, Out-of-Network Referrals and Waiver of Patient Copayments and Deductibles: The Battle Between Payors and Providers Endures and Intensifies, Health Law., June 2013, at 18, 20 (describing “pull- through” schemes as method of influencing physicians and possibly violative of federal antikickback law). 209. The leading example is a large settlement with Hospital Corporation of America (HCA), $225.5 million of which was for allegations that HCA provided physicians with various inducements to increase the number of Medicare referrals to HCA facilities. See Press Release, U.S. Dep’t of Justice, Largest Health Care Fraud Case in U.S. History Settled: HCA Investigation Nets Record Total of $1.7 Billion (June 26, 2003), http://www.justice.gov/opa/pr/2003/June/03_civ_386.htm (on file with the Columbia Law Review). Many other large FCA settlements have included kickback or self-referral claims, including two noted previously: the $3 billion settlement with Glaxo, which included allegations that Glaxo paid kickbacks to physicians to prescribe drugs for noncovered uses, see supra note 5 and accompanying text, and the $2 billion settlement with J&J, which included allegations that J&J induced Omnicare, the nation’s largest drug dispenser in nursing homes, to purchase and then sell J&J’s drugs to nursing-home patients for treatment of unapproved uses, see supra note 1 and accompanying text. Smaller examples include a recent settlement in which Orthofix, a medical-device company, paid $42 million to settle allegations that it paid kickbacks to induce sales of the company’s bone-growth stimulators, see Press Release, U.S. Dep’t of Justice, Orthofix, Inc. Agrees to Plead Guilty to Felony (June 7, 2012), http://www.justice.gov/usao/ma/news/ 2012/June/OrthofixSettlementPR.html (on file with the Columbia Law Review), and also a $161 million settlement with Caremark that included allegations that the company paid kickbacks in its home-infusion, oncology, hemophilia, and human-growth-hormone businesses, see Press Release, U.S. Dep’t of Justice, Caremark to Pay $161 Million in Fraud and Kickback Cases (June 16, 1995), http://www.justice.gov/opa/pr/Pre_96/June95/342. txt.html (on file with the Columbia Law Review). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1983 rendered necessary and properly administered medical services, because they had certified compliance, whether expressly or impliedly, with all governing laws.210 As with certification claims more generally, qui tam suits asserting kickback claims thus depend on relators’ ability to prove that government payments were conditioned upon compliance with other, often separate statutory or regulatory mandates. More importantly, FCA kickback claims were jurisdictionally innovative because they “hitched” the FCA to federal kickback and self-referral laws despite congressional creation of a separate, purely public enforcement mechanism that did not include a private right of action among the options.211 For instance, at the time of the first wave of qui tam suits asserting kickback claims in 1994 and 1995,212 Congress had empowered the DOJ to bring criminal actions targeting kickback arrangements, but it had authorized the government’s public enforcers to seek civil monetary penalties for only a narrow band of kickback conduct, and it had plainly not authorized private litigants to seek such penalties at all.213 As a result, kickback cases are the poster child for the concern that qui tam suits can be used to colonize entirely separate regulatory regimes for which

210. See supra notes 120–121, 152–153 and accompanying text (defining and discussing rise of controversial certification claims). 211. As courts have recognized, the legislative history surrounding the kickback laws shows that Congress did not intend to create a private right of action. See W. Allis Mem’l Hosp., Inc. v. Bowen, 852 F.2d 251, 254–55 (7th Cir. 1988) (reviewing legislative history of Social Security Act amendments and finding no private right of action); Donovan v. Rothman, 106 F. Supp. 2d 513, 516 (S.D.N.Y. 2000) (finding no private right of action under antikickback statute). Rather, the statute grants the Office of Inspector General of the Department of Health and Human Services and the Federal Bureau of Investigation (FBI) primary enforcement responsibility. See 42 U.S.C. §§ 1320a-7a to -7b (2012) (providing civil-enforcement authority to Office of Inspector General, Department of Health and Human Services, and criminal-enforcement authority to FBI). 212. Relators filed at least five suits from 1994 to 1996. See Robert N. Rabecs, Kickbacks as False Claims: The Use of the Civil False Claims Act to Prosecute Violations of the Federal Health Care Program’s Anti-Kickback Statute, 2001 Mich. St. L. Rev. 1, 44–59 (discussing United States ex rel. Pogue v. American Healthcorp, Inc. (1996), United States ex rel. Roy v. Anthony (1994), United States ex rel. Parker v. Apria Healthcare Group. (1995), United States ex rel. Montagano v. Midway Hospital Medical Center (1995), and United States ex rel. Thompson v. Healthcare Corp. (1995)). 213. Congress expanded the AKS twice as the first qui tam suits got off the ground in the mid-1990s, increasing civil monetary penalties (including a $50,000 penalty for AKS violations) and making them available for a wider range of activities and programs, strengthening available administrative remedies, and elevating criminal infractions from a misdemeanor to a felony. See Balanced Budget Act of 1997, Pub. L. No. 105-33, §§ 4304(b), 4314, 4331(c), 111 Stat. 251, 383–84, 389, 396 (codified as amended at 42 U.S.C. §§ 1320a-7a, 1395nn(g), 1320a-7 (2012)). Under the Health Insurance Portability and Accountability Act of 1996, the AKS was expanded to cover all federal healthcare programs except the Federal Employee Health Benefit Program. See Health Insurance Portability and Accountability Act of 1996, Pub. L. No. 104-191, § 204, 110 Stat. 1936, 1999–2000 (codified as amended at 42 U.S.C. § 1320a-7b (2012)). 1984 COLUMBIA LAW REVIEW [Vol. 114:1913

Congress has prescribed an alternate enforcement scheme that does not include a private right of action.214 Second, the kickback and self-referral area adds unique analytic value because it promises a better-developed legislative and administrative record than the AWP or royalty-valuation areas from which one might draw inferences about the degree of political control within the system. For starters, kickback suits offer a rare case study in which a principal legal question raised by the initial qui tam suits—the jurisdictional question, just noted, as to whether the Anti-Kickback Statute (AKS) preempts or otherwise precludes FCA kickback actions—ultimately saw decisive congressional resolution. After more than a decade of lower-court debate on the issue, Congress formally authorized qui tam kickback actions in 2009 as part of the Patient Protection and Affordable Care Act’s (PPACA) overhaul of the nation’s healthcare system by making any claim for payment while in violation of the AKS a “false claim” under the FCA.215 In theory, at least, one can thus track legislative efforts leading up to congressional sanction of qui tam kickback claims as a way to assess the effect of private enforcement on legal development. Just as important, qui tam suits asserting kickback claims have gone forward against the backdrop of rules promulgated by the OIG of the Department of Health and Human Services exempting certain practices from kickback liability.216 Because these “safe harbor” regulations immu-

214. See supra notes 9, 118–121 (cataloguing these claims in scholarly and journalistic press); see also John T. Boese & Beth C. McClain, Why Thompson Is Wrong: Misuse of the False Claims Act to Enforce the Anti-Kickback Act, 51 Ala. L. Rev. 1, 49 (1999) (decrying possibility qui tam relator can “trump[]” government’s discretion not to prosecute under federal antikickback laws). A possible response to these views is that the legislative history to the 1986 FCA amendments specifically notes, “a false claim may take many forms, the most common being a claim for goods or services . . . provided in violation of . .. [a] statute[] or regulation.” S. Rep. No. 99-345, at 9 (1986). 215. See 42 U.S.C. § 1320a-7b(g) (2012). On the judicial debate about whether AKS violations can serve as a predicate for FCA liability, see Hanna Lundqvist et al., Health Care Fraud, 49 Am. Crim. L. Rev. 863, 925 n.450 (quoting Mark Allen Kleiman who asserts “judicial consensus is slowly building that illegal kickbacks are false claims”); Rabecs, supra note 212, at 43–59 (collecting and analyzing district court cases). Until Congress resolved the question in 2009, the two leading decisions were United States ex rel. Thompson v. Columbia/HCA Healthcare Corp., 125 F.3d 899, 902–03 (5th Cir. 1997), which allowed the government’s argument that the defendant’s alleged violations of the AKS constituted a false claim, and United States ex rel. Pogue v. American Healthcorp, Inc., 914 F. Supp. 1507, 1509–10 (M.D. Tenn. 1996), which held that a kickback violation could support an FCA violation. For early decisions on whether the AKS might preempt even public enforcement actions under the FCA, see United States v. Gen. Dynamics Corp., 19 F.3d 770, 776–77 (2d Cir. 1994) (holding AKS does not preempt remedies of United States under FCA); United States v. Killough, 848 F.2d 1523, 1526–27 (11th Cir. 1988) (illustrating additional civil remedies despite AKS claim). 216. Congress first directed the Department of Health and Human Services to create safe harbors via the Medicare and Medicaid Patient and Program Protection Act, Pub. L. No. 100-93, § 14(a), 101 Stat. 697, 697 (1987), reprinted in 42 U.S.C. § 1320a-7b app. (1994). The initial safe harbors were promulgated in the early 1990s, then significantly 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1985 nize defendants from FCA liability as well,217 the result is an extensive administrative record across numerous rulemakings from which one might assess how, if at all, qui tam enforcement efforts have altered the path of legal development by reshaping the political environment. Given this richer legislative and administrative record, what further inferences can be drawn about the degree of political control within the FCA regime? To begin, there is once more strong evidence that qui tam enforcement efforts have exploited political—and, in particular, administrative—inertia. Most damning in this regard is the fact that, at the time of the first qui tam self-referral suits under the new Stark Law, prosecutors and agency administrators had not yet initiated any public enforcement efforts of their own.218 The main reason, commentators have suggested, is that HCFA’s rulemaking process had suffered substantial delays that would ultimately push promulgation of final rules to six years after the law’s effective date, and agency administrators and federal prosecutors thought it unfair to bring enforcement actions until implementing regulations defining key statutory terms were in place.219 Like the AWP and royalty-valuation cases, the initial qui tam self-referral suits thus plainly pushed into an area of substantial regulatory ambiguity, this time as a slow-moving administrative apparatus strained to flesh out key regulatory provisions. Beyond this, the legislative and administrative record once more helps to sketch a range of interpretive possibilities but permits few firm conclusions about the overall degree of political control within the system. Focusing first on the administrative side of the equation, it is important to note that the DOJ had, by 1994, initiated (and settled) its own lawsuit—that is, brought an FCA action that did not originate with a qui tam action—in a case that sought to hitch the FCA to AKS viola-

revised in the later 1990s. See 42 C.F.R. § 1001.952 (2013) (listing safe harbors); see also Tim Drake, Alexandra Kanu, & Nick Silverman, Health Care Fraud, 50 Am. Crim. L. Rev. 1131, 1149–67 (2013) (cataloguing safe harbors). 217. Most courts have found that safe-harbor compliance precludes FCA liability by negating the FCA’s intent element. See, e.g., United States ex rel. Bott v. Silicon Valley Colls., 262 F. App’x 810, 812 (9th Cir. 2008) (“If defendants complied with a facially valid regulation, relators cannot show the required scienter under the False Claims Act for actions after the safe harbor regulation was promulgated.”). Safe harbors can also create qui tam liability where none existed before. See, e.g., United States ex rel. Streck v. Allergan, Inc., 894 F. Supp. 2d 584, 605 (E.D. Pa. 2012) (granting motion to dismiss qui tam action for conduct prior to 2007 promulgation of safe-harbor rule but denying motion for post-2007 conduct). 218. In the HCA cases, as described infra note 223, various relators asserted Stark self-referral claims in addition to AKS kickback claims. 219. See Krause, Regulating, supra note 9, at 249–52 (providing detailed overview of chronic delays in rulemakings regarding Stark self-referral laws); Salcido, Unleashes, supra note 9, at 1, 3 (noting belief among qui tam defense counsel that prosecutors were awaiting “final regulations interpreting the full scope of the Stark Law”). 1986 COLUMBIA LAW REVIEW [Vol. 114:1913 tions.220 And by 1996, the DOJ had, having declined to join two prior qui tam cases, intervened in a case asserting FCA claims predicated on kickback violations.221 Such evidence is inconsistent with a claim that qui tam suits asserting kickback claims generated a type of enforcement that would not have existed at all but for the presence of a private enforcement mechanism. The interpretations that remain point in opposite directions regarding the degree of political control within the system. One possibility is that qui tam enforcement efforts, by forcing an early administrative reckoning with the propriety of using FCA suits to enforce kickback laws, critically impacted the political process by shortchanging full interagency ventilation of the issue or by forcing the government into a premature position without the benefit of formal input via the notice-and-comment- rulemaking process. On this view, the presence of a qui tam enforcement mechanism altered political dynamics by compressing and thus potentially distorting the debate.222 But it is also important to recall that, as with the AWP suits, the first large kickback settlements, particularly the first several hundred million dollars in settlements with the Hospital Corporation of America (HCA) in 2001, came only after the DOJ had agreed to intervene in the qui tam action.223 To that extent, a counter- interpretation seems just as plausible: The DOJ’s statutory gatekeeper powers kept it firmly in the driver’s seat throughout, but the agency was also resource constrained and risk-averse and so preferred to rely upon

220. See Rabecs, supra note 212, at 43–44 (discussing Shalala v. T2 Medical Inc., where DOJ asserted infusion-therapy center violated FCA via illegal kickback referrals). However, some have contested the significance of Shalala in this regard because the case was brought by the DOJ Criminal Division and asserted criminal and civil claims. See Robert Salcido, Mixing Oil and Water: The Government’s Mistaken Use of the Medicare Anti-Kickback Statute in False Claims Act Prosecutions, 6 Annals Health L. 105, 124–25 (1997) (“It was unclear whether the government had asserted, in fact, that it may predicate a civil FCA action upon a violation of the Anti-Kickback Act or whether the criminal investigation had unearthed alleged violations of the Anti-Kickback Act while the civil investigation had uncovered allegations of unrelated violations of the FCA.”). 221. See United States ex rel. Parker v. Apria Healthcare Grp., No. 1:95-cv-02142 (N.D. Ga. filed Aug. 23, 1995) (reflecting notice to intervene in part and to decline to intervene in part by United States filed June 27, 1996). The two cases in which the DOJ declined to intervene were Thompson and Pogue, as described infra note 223. 222. See David A. Hyman, Health Care Fraud and Abuse: Market Change, Social Norms, and the Trust “Reposed in the Workmen,” 30 J. Legal Stud. 531, 565 (2001) (“The availability of qui tam proceedings has also influenced enforcement priorities.”). 223. See United States ex rel. Parker v. Apria Healthcare Grp., No. 1:95-cv-02142- FMH (N.D. Ga. filed Aug. 23, 1995) (reflecting notice to intervene in part and to decline to intervene in part by United States filed June 27, 1996, and otherwise informing court of parties’ settlement filed Dec. 23, 1996); see also United States ex rel. Thompson v. Columbia/HCA Healthcare Corp. of Cent. Tex., No. 2:95-cv-00110 (S.D. Tex. filed Mar. 10, 1995) (reflecting advisory memorandum by United States filed Aug. 4, 1995, and closure of case on Jan. 10, 2000); United States ex rel. Pogue v. Am. Healthcorp, Inc., No. 3:94-cv-00515 (M.D. Tenn. filed June 23, 1994). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1987 private enforcers to test the waters in federal court before diving in and spending the agency’s reputational capital and resources. Nor does the voluminous administrative record around various “safe harbor” rulemakings offer any clear and concrete evidence that qui tam enforcement efforts broke apart legislative coalitions or otherwise reshaped the political environment in ways that pushed the substance of the law in a different direction than it would have traveled in the absence of qui tam suits. Perhaps the best opportunity to identify evidence of a feedback effect of the sort Part I theorized is OIG’s 1999 promulgation of rules clarifying the safe harbors first put into place back in 1994, before the first qui tam suits asserting kickback violations were filed.224 Those clarifications made a number of changes to the previous safe-harbor regulations in areas that had been the subject of early qui tam enforcement efforts asserting kickback claims, including provisions governing physician-investment interests225 and the practice of “bundling” together medical goods or services in sales between entities.226 The 1999 clarifications also contained telling omissions. The rules did not, despite substantial industry pressure, create a safe harbor on another issue that had begun to draw qui tam enforcement activity: the propriety of “integrated delivery systems,” including referrals between distinct departments of a single healthcare entity or between an entity and a wholly owned subsidiary.227 However, consideration of the administrative record once more does not reveal any concrete evidence of cleavages within industry on the reach of the kickback laws that evinces a feedback effect of the sort Part I

224. For the final rules, see Safe Harbor Provisions Under Health Care Anti-Kickback Statute, 64 Fed. Reg. 63,518 (Nov. 19, 1999) (codified at 42 C.F.R. pt. 1001). For the Notice of Proposed Rulemaking, see Clarification of the OIG Safe Harbor Anti-Kickback Provisions, 59 Fed. Reg. 37,202 (July 21, 1994) (codified at 42 C.F.R. pt. 1001). 225. See 42 C.F.R. § 1001.952(a) (2013) (listing safe-harbor exemptions from kickback liability surrounding investment interests). A key provision of the 1999 regulations substantially widened liability for failure to meet the rule that no more than forty percent of investment interests be held by “tainted” investors in a position to make referrals by making clear that a wider range of actors other than physicians, including hospitals, nursing homes, and managed-care companies, qualified as “tainted” investors within the meaning of the rule. See Safe Harbor Provisions Under Health Care Anti- Kickback Statute, 64 Fed. Reg. at 63,523. The new rules also prohibited a healthcare entity or its investors from loaning or guaranteeing loans for a “tainted” investor who uses the loan to obtain an ownership interest in the entity. See Linda A. Baumann, Navigating the New Safe Harbors to the Anti-Kickback Statute, Health Law., Feb. 2000, at 1, 4–6, 8 (describing modifications to liability for investment interests and loans for “tainted” investors). 226. See 42 C.F.R. § 1001.952(h) (2013) (listing safe-harbor exemptions from kickback liability surrounding discounts); Baumann, supra note 225, at 8 (describing expanded protections for “bundled” sales arrangements in 1999 regulations). 227. See Baumann, supra note 225, at 4 (describing OIG’s denial of safe harbor for integrated delivery systems or wholly owned subsidiaries); Drake, Kanu & Silverman, supra note 216, at 1167 & n.240 (describing proposed safe harbor for integrated delivery systems and wholly owned subsidiaries not included in final 1999 rules). 1988 COLUMBIA LAW REVIEW [Vol. 114:1913 hypothesized. Indeed, the Office of the Inspector General at HHS, in promulgating its 1999 rule adding safe harbors and clarifying existing ones under the AKS, offered over thirty pages of analysis of the public comments received in response to its notice of proposed rulemaking but nowhere suggested that industry groups or specific industry members were divided on whether safe harbors should be expanded or narrowed in their coverage.228 This lack of concrete evidence of a feedback effect is, as with the AWP example, perhaps not surprising. Particularly noteworthy is the fact that the 1999 regulations predated the first major kickback settlements. And while leading industry actors like HCA were in heated litigation of qui tam suits filed before the 1999 regulations, none of those industry actors had yet suffered an adverse judgment or concluded a settlement, making it less than certain that political action would be necessary at all during the run-up to the 1999 rules.229 Nor, as noted previously, does a lack of concrete evidence in the publicly available record around those rules foreclose the possibility of a feedback effect, as lobbying can—and, indeed, might be more likely to—occur behind the scenes.230 Still, the kickback and self-referral area constitutes a second major flavor of qui tam activity—with the AWP suits being the other—for which the record does not contain any concrete evidence of Part I’s hypothesized feedback effect.231

228. See Safe Harbor Provisions Under Health Care Anti-Kickback Statute, 64 Fed. Reg. at 63,518–51. 229. It is also notable that the 1999 regulations mostly liberalized the practices that qualified for safe-harbor protection. But the feedback effect and increasing-returns process that Part I identified is more observable where political action expands the reach of a law or regulation so that a litigation loser’s competitors are exposed to the same liability that it was. Where, by contrast, the regulatory process liberalizes prohibitions, the analyst must further identify instances in which regulations were liberalized less than they otherwise might have been. 230. See supra note 181 and accompanying text (highlighting weak constraints on ex parte lobbying in legislative process). 231. Still another explanation for the lack of concrete evidence of Part I’s hypothesized feedback effect is that kickback liability did not induce qui tam defendants who suffered adverse judgments or entered into settlements to make changes in their operations or business models that would be difficult to reverse later. See supra note 95 and accompanying text (discussing importance of this characteristic to plausibility of feedback or increasing-returns mechanism). That is, unlike in the AWP context, where early litigation losers entered into CIAs compelling them to utilize a particular reimbursement methodology going forward, a change in the law regarding the types of investment-based remuneration that a healthcare entity can pay to physicians in a position to refer patients to the entity might not be irreversible, as an entity that had suffered an early judgment and ceased paying such remuneration could merely resume doing so if the rulemaking process subsequently provided safe-harbor protection. However, many forms of conduct that implicate federal kickback and self-referral laws are, in fact, difficult to reverse because they involve complex organizational structures and contractual relationships that are difficult to unwind. And indeed, OIG fielded substantial industry requests during the 1999 rulemakings to provide for a transition period, with many 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1989

Finally, moving from the administrative to the legislative record fares no better in generating firm conclusions about the nature of political control within the FCA regime. The heated hearings leading up to Congress’s enactment of the False Claims Act Correction Act of 2008 offer a representative example, with the healthcare industry presenting a united front on “certification” theories of FCA liability via sharply word- ed testimony by an attorney representing the Chamber of Commerce who also formerly represented many healthcare companies in qui tam litigation.232 Among other things, the Chamber bitterly complained of qui tam relators’ efforts to “bootstrap regulatory violations onto the FCA.”233 But the hearings offer no hint of cleavages among industry actors on kickbacks and self-referrals—or, for that matter, any other aspect of the FCA regime. The only voice at the hearings that dissents from the Chamber’s testimony is a former federal prosecutor and member of the qui tam plaintiffs’ bar.234 Much the same was true of testimony nearly a decade earlier during hearings focused not on the FCA but rather on Medicare “self-referral laws” more generally. Appearing at these hearings was a somewhat wider range of industry groups, with several offering testimony that was harshly critical of mounting qui tam liability.235 But there is no indication of substantive disagreement among industry actors on the reach of the self-referral laws. Though behind-the-scenes lobbying remains possible, there is once more no direct evidence that qui tam suits altered the political equation by inducing litigation losers to lobby in favor of an unforgiving liability standard going forward. Still, the manner in which legislative and administrative activity in the kickback and self-referral context has played out, though falling well industry actors noting the difficulty of unwinding such arrangements upon a change in governing law. See Safe Harbor Provisions Under the Anti-Kickback Statute, 64 Fed. Reg. 63,518, 63,520 (Nov. 19, 1999) (codified at 42 C.F.R. pt. 1001) (noting commenters’ requests for “reasonable grace period to restructure existing arrangements” where new regulations altered requirements); see also Drake, Kanu & Silverman, supra note 216, at 1167 & n.241 (describing industry requests for transition period and OIG rejection of those requests). 232. See The False Claims Act Correction Act (S. 2041): Strengthening the Government’s Most Effective Tool Against Fraud for the 21st Century: Hearing on S. 2041 Before the S. Comm. on the Judiciary, 110th Cong. 75, 82 (2008) (statement of John T. Boese) (noting false certification as basis for liability). 233. Id. at 82–83 & n.4. 234. Id. at 136 (statement of John E. Clark) (explaining support for proposed amendments in bill). 235. See Medicare Self-Referral Laws: Hearing Before the Subcomm. on Health of the H. Comm. on Ways & Means, 106th Cong. 121 (1999) (statement of the Federation of American Health Systems) (noting “harsh result” of qui tam liability for minor infractions under Stark Laws); id. at 58 (statement of Sanford V. Teplitsky, Past President, National Health Lawyers Association) (noting “ambiguous and confusing” nature of Stark Laws and proposed implementing regulations and expressing concern about “qui tam actions by private litigants” given that ambiguity). 1990 COLUMBIA LAW REVIEW [Vol. 114:1913 short of a clear-and-convincing showing of a feedback effect, offers some critical insights about the degree of political control within the FCA regime. For one, Congress’s sluggish embrace of using the FCA to enforce the AKS, after nearly twenty years of litigation over the propriety of doing so, fortifies Part I’s concern about the limited legislative will and capacity to exercise meaningful oversight of statutory elaboration driven by litigation. A plausible explanation for Congress’s belated response is that rational legislators saw little political upside, and perhaps even substantial downside, to entering the fray and addressing the FCA’s use in the kickback area, particularly if judicial interpretation seemed headed in the right direction. This was especially true as many lower courts imposed a “materiality” requirement that quelled the worst fears of commentators that “certification” theories of FCA liability would extend to any and all statutory or regulatory violations, no matter how divorced from the federal government’s payment of money.236 Perhaps, then, one should not be surprised that Congress was more rearguard than vanguard in embracing an FCA role in enforcement of the AKS. Still, Congress’s performance hardly inspires confidence in the legislative will or capacity to police statutory elaboration more generally. Two decades’ worth of administrative efforts to regulate kickbacks and self-referrals likewise inspire skepticism that private enforcement efforts can be policed via the ponderous rulemaking process. Exhibit A, of course, is HCFA’s protracted, multiyear rulemaking effort aimed at implementing the Stark Laws as qui tam suits sprouted all around.237 Yet even OIG’s continuing efforts to elaborate a network of safe harbors under the AKS have drawn criticism, with some commentators question- ing the efficacy, and even the relevance, of the safe-harbor approach in a fast-moving, dynamic industry like healthcare.238 On this view, administrative rulemaking is barely an ex ante regulatory device at all but rather functions as an ex post, and often irrelevant, response to conduct that has long since been tested via public or private enforcement efforts. Put another way, regulatory efforts are more akin to the carnival game of “whack-a-mole,” in which the creaky rulemaking process perennially plays catch-up, than a deliberate, ex ante enterprise. When combined with congressional sluggishness in embracing FCA kickback enforcement, the rulemaking experience in the kickback and self-referral space yields a broader, and final, observation: Frontline control over the qui tam regime is unlikely to depend solely—or perhaps even partly—on congressional lawmaking or administrative rulemakings. Rather, it is the DOJ, in its exercise of its case-by-case gatekeeper duties

236. See supra notes 153, 161 and accompanying text (discussing “materiality” requirement). 237. See supra note 219 and accompanying text (noting and explaining delays in HCFA rulemaking). 238. See Matthew, Tainted, supra note 124, at 551–52 (“[Safe harbors] work at cross- purposes with facilitating a financially healthy health care market.”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1991 the FCA vests in it, that seems best positioned to police private legal innovation and statutory elaboration within the system. As noted previously, this fact highlights the need for a more systematic accounting of DOJ oversight in evaluating critics’ claims that the current FCA regime is “abandoning public enforcement of the antifraud statutes”239 or otherwise suffers from a lack of political control. The next and final section of this Part now turns to that task.

C. DOJ Oversight: Delegation or Decoupling? Figure 7’s analysis charting a rise in multidefendant actions toward the start of the post-1986 regime, when combined with the above case studies of several marquee flavors of qui tam litigation, offers a window into private legal innovation within the FCA regime. However, the analysis provides only suggestive evidence regarding the overall degree of public control of qui tam litigation efforts. This final section, while continuing to acknowledge the difficulty of drawing clean inferences on the public-control issue, seeks to make some further progress by focusing attention on the gatekeeper role played by the DOJ as the institution best positioned to police legislative fidelity within the system. At one level, this inquiry might seem an easy one, as prior empirical work suggests that the DOJ exercises firm control over qui tam enforcement efforts using the case-by-case gatekeeper powers the FCA vests in it. The best evidence in this regard is what might be called the 90/10 effect: Since 1986, roughly ninety percent of qui tam suits in which the DOJ exercises its statutory authority under the FCA to intervene in and take over control of the litigation go on to achieve a positive-dollar recovery, while only ten percent of suits in which the DOJ declines to intervene do so.240 To be sure, it is difficult to determine whether this is because of the DOJ’s ability to screen meritorious cases (a “selection” effect) or instead reflects the DOJ’s unique litigation leverage (a “treatment” effect).241 But whatever the mechanism, the end result is the same: Only a small percentage of relators achieve a recovery where the DOJ is not on board. To that extent, perhaps one should be less concerned about the DOJ’s capacity to police legislative fidelity. There is, however, still plenty of room for skepticism about the DOJ’s willingness and capacity to police legislative fidelity. The main concern has already drawn mention: The DOJ’s role within the system implicates complex separation-of-powers and interagency dynamics, including the possibility that DOJ decisionmaking is a prosecutorial end- run that deviates from, rather than implements, legislative bargains.

239. Id. at 566–67. 240. See Engstrom, Public Regulation, supra note 85, at 1720–22 (presenting data on 90/10 point). 241. Id. at 1737 (noting difficulty of prying apart selection and treatment effects in agency-gatekeeper context). 1992 COLUMBIA LAW REVIEW [Vol. 114:1913

Anecdotal claims along these lines abound: Defense-side critics have long complained that the DOJ’s intervention decisionmaking is overly focused on large, marquee recoveries as a way to burnish the agency’s reputation with the public and keep congressional overseers at bay.242 Relator-side critics, too, complain that the DOJ sometimes intervenes in cases only belatedly, on the eve of a settlement, suggesting that face saving, rather than careful deliberation, may be driving decisions.243 Prior empirical research offers further support, suggesting that DOJ decisions are in part strategic, driven by DOJ resource constraints and the identity (and corporate power) of the defendant.244 Taken together, this mix of evidence suggests that the DOJ does not place much of a premium on policing legislative fidelity at all—and, indeed, may exhibit something akin to the profit motivation that is presumed to drive private enforcers.245 Moreover, even if the DOJ maintains a firm hand over qui tam cases via its intervention authority, the DOJ’s exercise of its other statutory gatekeeper powers inspires less confidence. For instance, the data con- firm critics’ complaints that the DOJ only rarely deploys its authority to terminate qui tam cases outright, thus virtually ignoring the most direct means of policing undesirable private legal innovations.246 To that extent, prior empirical work strongly suggests, as Part I theorized, that DOJ oversight may suffer from a type of “bailout effect” because of the agency’s ability to shift the cost of case termination, both actual and reputational, to the courts.247

242. See id. at 1714 (collecting sources criticizing DOJ on these grounds). 243. See Fact Sheet: Examples of Cases in Which the Department of Justice Intervened After Substantial Time or Commitment of Resources by Whistleblowers and Their Attorneys, Taxpayers Against Fraud, http://www.taf.org/FactSheetB.pdf (on file with the Columbia Law Review) (last visited Oct. 8, 2014) (listing cases with delayed DOJ intervention). 244. See Engstrom, Public Regulation, supra note 85, at 1728–37 (analyzing “strategic” factors in DOJ decisionmaking). 245. See Lemos & Minzner, supra note 18, at 912–13 (“[F]inancially motivated public enforcers will tend to behave more like private enforcers.”). 246. See Engstrom, Public Regulation, supra note 85, at 1717–18 (finding DOJ exercises its termination authority in only negligible percentage of cases). This should perhaps not be surprising, as DOJ officials have repeatedly suggested during congressional testimony that the DOJ has “little inclination to put scarce public enforcement resources toward dismissing meritless cases over other, more affirmative enforcement efforts.” Id. at 1712 & n.70 (collecting testimony). On critics’ claims, see Matthew, Moral Hazard, supra note 124, at 301 (“[I]n the face of weak monitoring incentives, the Government will allow cases based on weak facts or even unfounded or experimental theories of recovery to proceed.”); Rich, supra note 8, at 1236–37 (“DOJ’s failure to dismiss non-meritorious qui tam actions more frequently has resulted in 94% of non-intervened qui tam suits . . . recovering no funds. These suits have exacted a heavy toll on defendants, the judicial system, and even relators.”). 247. See supra note 87 and accompanying text (outlining this logic). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1993

FIGURE 8: MEAN DOJ INVESTIGATION TIME BY CASE-ELECTION YEAR AND QUI TAM CASE “TYPE,” 1986–2011

Perhaps most importantly, there is reason to believe that the DOJ’s control over the system may be loosening. For instance, prior empirical work shows a clear uptick in recoveries in unintervened cases.248 Simi- larly, recent empirical work, while finding a high correlation between the DOJ’s intervention decisions and case outcomes over the life of the post- 1986 regime, also suggests that the DOJ may increasingly be over- whelmed by growing qui tam case flows. Figure 8 presents some of the clearest evidence in this regard, showing a steady increase in the number of days that qui tam suits have remained under investigation prior to a DOJ case-election decision. Indeed, the data reveal that the average investigation time among qui tam cases that reached election decisions in 2011 was nearly two years, roughly twice the average investigation time during earlier periods. Rising investigation times may also be producing a reckoning of sorts with district court judges, who may be growing increasingly impatient with (and prone to deny) the DOJ’s serial requests for extensions of time to make intervention decisions.249 Some commentators

248. See Engstrom, Public Regulation, supra note 85, at 1721 fig.3, 1722 n.95 (noting declined actions yielding large settlements in 2007 and 2010). 249. See, e.g., United States ex rel. Martin v. Life Care Ctrs. of Am., Inc., 912 F. Supp. 2d 618, 623, 626 (E.D. Tenn. 2012) (describing length of time case was under seal as “absurd” and cautioning future government requests for extension “will be met with significant scrutiny”); United States ex rel. Underwood v. Genentech, Inc., No. 03-3983, 2010 WL 3955786, at *4 (E.D. Pa. Oct. 5, 2010) (“The Government has not explained why it needed six years to conduct its investigation. . . .”); United States ex rel. Dekort v. Integrated Coast Guard Sys., 705 F. Supp. 2d 519, 529 n.2 (N.D. Tex. 2010) (unsealing qui tam complaint and permitting relator to move forward after “grant[ing] the United States 1994 COLUMBIA LAW REVIEW [Vol. 114:1913 suggest growing judicial frustration is the main reason underpinning the DOJ’s relatively recent practice by which DOJ lawyers, rather than intervening or declining, inform the court that the government is “not intervening at this time.”250 The DOJ’s use of these “notices of no election,” as it terms them, appears to be substantial and increasing. Indeed, the data show that the DOJ has taken this position in at least 261 cases, beginning with only a handful of such elections in 2004 and growing to more than fifty per year in 2008 and 2009, the last years for which the right-censoring problem described previously permits a reasonably accurate estimate.251 This suggests that a nontrivial proportion of qui tam suits are advancing to litigation with something less than full DOJ review or participation. The broader challenge is linking this evidence of a shifting DOJ oversight role to the public-control questions at the core of Part I’s theoretical framework and, relatedly, critics’ claims about a lack of public control within the qui tam regime. One possible interpretation is that the FCA regime is decoupling from DOJ control as the regime shifts into a higher-volume, retail mode, and a resource-strapped DOJ struggles to maintain the quality of its case-screening efforts. More concretely, the uptick in declined-case recoveries together suggests that qui tam relators may be placing less and less trust in DOJ case screening, preferring to test case merit themselves rather than filing an immediate voluntary dismissal upon a DOJ declination. As prior research has shown, some relator-side firms have built reputations for doing just that, with one particularly active firm exhibiting a low DOJ intervention rate in qui tam cases it files compared to other repeat-play firms, but boasting one of the

numerous extensions of its deadline to intervene”). For earlier cases, see United States ex rel. LaCorte v. Smithkline Beecham Clinical Labs., Inc., No. CIV. A. 97-09412, 1998 WL 840012, at *2 (E.D. La. Nov. 12, 1998) (“The court has already . . . allow[ed] extensions, and the government has had ample time to consider whether it will intervene in the case . . . . The United States now seeks, two months after it promised to unseal the case, another extension . . . . The court takes this disingenuous behavior seriously.”); United States ex rel. Costa v. Baker & Taylor, Inc., 955 F. Supp. 1188, 1191–92 (N.D. Cal. 1997) (expressing “regret” at having granted earlier extensions and noting “relative ease of granting, rather than denying, these extensions may too often lead courts to prolong unnecessarily the period of the seal”). 250. See Inst. for Legal Reform, U.S. Chamber of Commerce, The New Lawsuit Ecosystem: Trends, Targets and Players 68 (2013), available at http://www.institutefor legalreform.com/uploads/sites/1/web-The_New-Lawsuit-Ecosystem-Report-Oct2013_2. pdf (on file with the Columbia Law Review) (“As courts lose patience with requests for extensions, DOJ is consenting to unsealing cases, allowing the qui tam plaintiff to proceed alone, with a notice that it will not intervene ‘at this time.’”). For an example of a case in which the DOJ used a “not at this time” case election, see Notice of the United States that It Is Not Intervening at This Time, United States ex rel. Shea v. Verizon Bus. Networks Servs., 904 F. Supp. 2d 28 (D.D.C. 2012) (No. 1:09-CV1050 GK). 251. These figures are derived from the DOJ-provided data by tallying up the number of entries in which the DOJ reported filing a “Notice of No Election” in the data field titled “Election Decision.” 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1995 highest ultimate recovery rates.252 If more relators and relator-side firms follow this path based on a perception that the DOJ’s case-screening acumen is eroding or that it simply lacks the resources to investigate cases effectively within the timeframes allowed by increasingly impatient courts, there may be greater space for private legal innovation going forward, and relatively less public control. But other interpretations are also possible. For instance, the above evidence is just as consistent with smarter and more efficient DOJ delegation of prosecutorial authority as it is with systemic decoupling. In particular, the rapid emergence of a highly specialized relators’ bar has given the DOJ access to a ready pool of repeat-play private enforcers with strong reputational incentives to toe the government line and predict, rather than force, agency enforcement priorities. Delegation to this group of specialized enforcers via nonintervention may thus present a lower risk of statutory “drift.” Far from suggesting a loss of DOJ control, the rise in unintervened recoveries might instead fuel the conclusion that the system is moving closer to something like the optimum imagined by private enforcement’s champions, in which the DOJ selectively dele- gates enforcement authority to competent and trustworthy private enforcers but remains firmly in control of macrolevel enforcement priorities.253 A final, and related, interpretive ambiguity is that it remains to be seen whether declining retail-level control over enforcement efforts— whether because of an erosion of the DOJ’s case-screening capacity or a conscious increase in delegation—translates into a declining DOJ capacity to identify and police entirely new theories of FCA liability at the wholesale level. Indeed, even a resource-strapped DOJ that is unable to fully screen or monitor the fiftieth Medicaid upcoding case within the regime may still be fully capable of recognizing novel liability theories when they arise and devoting whatever time and energy is necessary to come to ground on their likely regulatory value and fidelity to legislative purposes. Put another way, a decline in micro-, case-level control over individual qui tam enforcement efforts need not imply a macrolevel loss of control over legal development. In the end, the empirical evidence suggests a changing DOJ oversight role, but it is difficult to determine whether the change is, on balance, positive or negative—or even if the shift is toward decoupling, delegation, or a mix of the two.

** *

This Part’s quantitative and qualitative analyses round out the understanding of the pathways down which qui tam litigation has traveled by

252. See Engstrom, Harnessing, supra note 18, at 1305 (collecting DOJ intervention and recovery rates for various firms). 253. See Engstrom, Public Regulation, supra note 85, at 1714 (discussing theories of DOJ’s welfare-maximizing role). 1996 COLUMBIA LAW REVIEW [Vol. 114:1913 confronting, as best as possible, the political-control concerns voiced by qui tam’s critics and at the core of Part I’s theoretical framework. In contrast to the quantitative evidence presented in Part III, the above analysis plainly falls short of a definitive showing that qui tam litigation has or has not evaded meaningful public control. Even so, this evidence—including suggestive evidence regarding increases in the scale and scope of qui tam litigation efforts since 1986, the clear tendency of several of the more significant flavors of qui tam enforcement to fill statutory and regulatory gaps left by political inertia, and a seeming shift in the DOJ’s gatekeeper role—paints a more worrying portrait than Part III’s firm rejection of widespread claims about a qui tam “explosion.” Change is clearly afoot within the FCA regime, and the regime may even be at something of a crossroads. To that extent, the above analyses should be seen as a beginning, not a culmination, of an inquiry focused on determining where, precisely, the regime might be heading from here.

V. PRIVATE ENFORCEMENT’S PATHWAYS REVISITED

This Article has explored private enforcement’s pathways by con- structing a novel theoretical framework and then offering an initial empirical test of key pieces of the theory via empirical analysis of qui tam litigation since 1986. This final Part steps back and briefly places the theory and findings in context and sketches some of their implications for debate about how, if at all, to revise the FCA; the choice between public and private enforcement in regulatory design more generally; and the relationship between litigation and democracy.

A. FCA Implications The above analysis contributes to current legal and policy debate about the FCA itself. To begin, Part III’s anatomy of qui tam’s post-1986 growth permits rejection of some of the more tendentious claims made by qui tam’s critics.254 Simply put, there is little evidence that the regime has experienced (or is experiencing) “explosive” growth or that qui tam enforcers are engaged in a “gold rush” of enforcement effort.255 Rather, the regime’s growth is best characterized as steady maturation.256 Critics should stop suggesting otherwise—or at least should more clearly specify key terms in characterizing qui tam litigation trends.

254. See supra notes 8, 115–116 (collecting claims regarding inefficient “explosion”). 255. See supra notes 8, 115–116 (cataloguing critics’ claim that qui tam regime is in midst of socially costly “explosion”). 256. See supra Part III.A (concluding empirical evidence “tends to undermine widespread characterizations of a qui tam ‘explosion.’”). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1997

The lack of evidence of a “gold rush” of qui tam enforcement activity likewise undercuts calls to cap relator awards.257 A principal rationale accompanying such calls is that large bounties incentivize inefficient speculation in lawsuits by relators and their counsel seeking “jackpot” payouts.258 Part III’s findings—particularly the finding that qui tam’s per- filing “efficiency” in returning funds to the federal fisc has not appre- ciably declined since 1986, even as high-dollar settlements have grown and drawn publicity259—plainly conflicts with such claims. A final set of implications focuses more squarely on issues relating to changes in the character of qui tam enforcement efforts and accompanying political-control concerns. For example, the evidence presented here offers an empirical baseline against which to evaluate a range of FCA reform proposals commentators have advanced in an effort to render legal innovation more transparent and strengthen the DOJ’s role in policing it. One such proposal would arm the DOJ, rather than agencies with primary responsibility within a given regulatory area, with rulemaking authority to define the bounds of FCA liability.260 Another would

257. See 155 Cong. Rec. 10,277–78, 10,508–13, 10,516 (2009) (discussing various amendments to FERA that would cap relator awards); id. at 10,517 (recording Senate roll- call vote rejecting proposed amendment capping relator awards at $50 million); see also Peter B. Hutt II & Anna R. Dolinsky, U.S. Chamber Inst. for Legal Reform, Preventing Government Overpayments to Qui Tam Plaintiffs: Proposed Amendments to the False Claims Act (2011), available at http://ilr.iwssites.com/uploads/sites/1/Proposed_ Amendments_to_the_False_Claims_Act.pdf (on file with the Columbia Law Review) (calling for caps on awards); Michael K. Loucks, Rewarding Whistleblowers Under the False Claims Act: The Great American Giveaway, Health Care Fraud Rep., Jan. 26, 2011, at 1, 1–2, available at https://www.skadden.com/sites/default/files/publications/Publications2338_ 0.pdf (on file with the Columbia Law Review) (same). 258. See Hutt & Dolinsky, supra note 257, at 14 (articulating this rationale). A second rationale is that large relator bounties are more than is necessary to overcome the professional and personal cost of whistleblowing and induce relators to come forward. See id. at 9–11 (calculating $15 million cap would leave whistleblowers $5 million after legal fees and taxes, offsetting lifetime lost earnings if retaliation ruins job prospects); see also Kesselheim et al., supra note 125, at 1838 (noting relator recoveries “appear[] to be quite disproportionate (in both positive and negative directions) to the whistle-blower’s personal investment in the case” and advocating “[m]ore sophisticated approaches to determining relators’ recoveries”); Loucks, supra note 257, at 3 (concluding caps would not deter relators from whistleblowing). This Article’s analysis does not directly address this second rationale. 259. See supra Figure 6 and accompanying text (showing qui tam case outcomes by filing year). 260. See Kovacic, Deterrent, supra note 136, at 237 (proposing DOJ be given rulemaking powers to define and delimit FCA liability theories); see also Joseph A. Grundfest, Disimplying Private Rights of Action Under the Federal Securities Laws: The Commission’s Authority, 107 Harv. L. Rev. 961, 1011–15 (1994) (proposing SEC be given similar power in securities law). Note that this proposal would upgrade the DOJ’s “gatekeeper” powers from a “retail” (i.e., case-by-case) to a “wholesale” (across-the-board) level and arguably give it greater authority to manage elaboration of the FCA’s antifraud mandate. See Engstrom, Gatekeepers, supra note 16, at 647–48 (coining this terminology for agency-gatekeeper powers). 1998 COLUMBIA LAW REVIEW [Vol. 114:1913 authorize courts in qui tam cases asserting novel liability theories to require the DOJ to provide a nonbinding certification that “the theory has legal merit and serves the public interest.”261 But Part IV’s analysis also counsels caution. In particular, the wisdom of adopting either of the above reforms will plainly turn on the diagnosis of the problem. Thus, for those who conclude from the above evidence that resource constraints are the primary threat to the DOJ’s ability to police private legal innovation, empowering district judges to impose further decision costs on the agency by ordering up the DOJ certification of FCA liability theories will merely stretch the DOJ’s limited resources even more thin than before. It is perhaps notable that the proposals seem mostly keyed to concerns about the institutional will of the DOJ to police private legal innovation, but do not address the DOJ’s capacity to do so. And yet, the evidence presented in Part IV.C on the DOJ’s exercise of its gatekeeper duties suggests that simple resource constraints may be the graver problem.262 To be sure, more work needs to be done in order to reach firmer conclusions about the degree of political control within the FCA regime and whether and how to tighten it. One potentially fruitful avenue is additional case studies exploring liability theories or industry sectors beyond those covered in Part IV.B.263 Another possibility is to focus on qui tam suits brought by competitor companies, rather than the more typical whistleblowing insiders, as a way to identify possible feedback effects.264 As noted repeatedly throughout Part IV, a reliable judgment on the public-control issue will come only with more, and more detailed, consideration of the complicated and dynamic interaction of private enforcers and judicial and political overseers.265

B. Regulatory Choice Beyond the FCA Other implications sweep well beyond the FCA. For instance, the theoretical and empirical analysis presented herein provides a richer understanding of the costs and benefits of deploying private litigation as

261. See Rich, supra note 8, at 1276–77. 262. See supra Figure 8 and accompanying text. 263. For instance, the interaction between agency rulemaking and qui tam suits appears to have been especially intense in the education space, with the Department of Education struggling to create safe-harbor regulations governing student-loan-recruiter compensation and then later voiding them. See United States v. Corinthian Colls., 655 F.3d 984, 989 n.1 (9th Cir. 2011) (describing DOE’s rulemaking efforts). 264. For example, qui tam settlements in the 1990s asserting various clinical laboratories had defrauded Medicare—sometimes called the “LabScam” cases—were filed by employees of a competitor laboratory as opposed to defendant-company insiders. See, e.g., Taxpayers Against Fraud, Settlements, False Claims Act & Qui Tam Q. Rev., Apr. 1997, at 30 (noting two relators in SmithKline Beecham Clinical Laboratories case “worked as sales personnel at SBCL competitors”). 265. See, e.g., supra note 253 and accompanying text (discussing DOJ’s evolving relationship with private enforcers). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 1999 a regulatory tool. As noted at turns above, much of the existing scholarly literature on enforcement suffers from at least two shortcomings. First, that literature has tended to focus on whether public or private enforcement delivers desired deterrence at lower social cost; to the extent the literature considers statutory “drift” or public control over legal development at all, it is modeled as a simple and abstract agency cost.266 This Article’s effort to more carefully specify the mechanisms by which private enforcement might uniquely drive legal development advances these scholarly efforts by facilitating empirical testing of key hypotheses and providing a foundation for thinking about the conditions under which drift is likely to arise or be significant. Second, existing efforts to model the use of private enforcement as a regulatory tool take a surprisingly static view.267 In contrast, the analysis presented here suggests that, as with other regulatory institutions,268 private enforcement regimes are highly dynamic in nature. Indeed, litigation regimes may follow a predictable evolutionary path or move through predictable “life cycles” of development, from early stages of relative inactivity before substantial private enforcement capacity is in place, to middle stages characterized by rising activity driven by publicity and private investment in enforcement infrastructure, and then to fully mature stages as the regime shifts into a full-scale retail mode. Impor- tantly, a litigation regime at each of these stages of development will generate very different regulatory outputs, with important implications for the degree of public control within the system. This richer, temporal account of private enforcement’s pathways expands, or at least reweights, the menu of factors that regulatory design- ers should consider in choosing between public and private enforcement modes (or the optimal mix of each). For instance, private enforcement may prove systematically less efficient than public enforcement due to more than the standard concern about the divergence of the social and private incentive to litigate.269 In addition, decentralized private enforcement may, relative to centralized public enforcement, equilibrate more slowly in response to shifting litigation incentives and other shocks to the litigation environment, including changes in governing law. Just as important, private enforcement may also generate substantial and undesirable statutory “drift” away from legislative purposes relative to purely public enforcement.

266. See supra notes 20, 56 and accompanying text (noting scholarly discussions of deterrence and social cost, but limited study of effects on legal development). 267. See supra note 20 and accompanying text (highlighting standard welfare- efficiency approach to studying private enforcement). 268. See Paul J. Quirk, Industry Influence in Federal Regulatory Agencies 14 (1981) (explaining regulatory agency “life cycle” based on predictable changes in political environment). 269. See supra notes 32, 49 and accompanying text (explaining that problems with private enforcement include equilibration costs). 2000 COLUMBIA LAW REVIEW [Vol. 114:1913

These types of insights are not merely academic. To the contrary, a richer understanding of private enforcement’s pathways can inform more specific regulatory-design debates. Perhaps the best example is the growing debate about how best to regulate securities fraud since the recent Dodd-Frank law substantially overhauled American financial regulation. Among other things, Dodd-Frank created a controversial whistleblower regime paying private individuals a bounty for providing the SEC with information leading to a successful public enforcement action.270 But Congress also ordered the SEC’s Inspector General to conduct a study to determine whether the current regime should be built out into a full-on qui tam regime by vesting whistleblowers with a private right of action.271 Published in early 2013, the Office of Inspector General’s report did not rule out a qui tam approach, noting the need for further study.272 And scholars have heeded that call, embarking on a number of efforts to systematically compare the costs and benefits of a “cash-for- information” and qui tam approach, focusing particular attention on either system’s ability to optimize the quantity and quality of whistleblower tips.273 The account of private enforcement’s pathways presented herein enriches this debate by providing a framework for weighing the broader efficiency and political-control trade-offs posed by the choice between the two approaches. On the one hand, the above analysis offers concrete evidence that a qui tam regime may, relative to a cash-for-information approach, prove more susceptible to statutory “drift” that is largely outside of public control as private enforcers rush to exploit regulatory

270. See Amanda M. Rose, Better Bounty Hunting: How the SEC’s New Whistleblower Program Changes the Securities Fraud Class Action Debate, 108 Nw. U. L. Rev. (forthcoming 2014) (manuscript at 22–23), available at http://ssrn.com/abstract_id= 2305403 (describing Dodd-Frank tip regime). 271. See Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 922(d), 124 Stat. 1376, 1848–49 (2010) (instructing Inspector General to study Dodd-Frank bounty regime, including possibility of “empowering whistleblowers or other individuals . . . to have a private right of action to bring suit based on the facts of the same case, on behalf of the Government and themselves, against persons who have committed securities fraud”). 272. See Office of Inspector Gen., SEC, Rep. No. 511, Evaluation of the SEC’s Whistleblower Program vi (2013), available at http://www.sec.gov/oig/reportspubs/511. pdf (on file with the Columbia Law Review) (“Upon collecting additional data and assessing the effectiveness of the program after a reasonable amount of time has passed, OIG will be in a better position to opine on the usefulness of adding a private right of action to the SEC’s whistleblower program.”). 273. See Anthony J. Casey & Anthony Niblett, Noise Reduction: The Screening Value of Qui Tam, 91 Wash. U. L. Rev. 1169, 1194–1207 (2013) (comparing efficiency of cash- for-information and qui tam bounty regimes in optimizing information quality and quantity); Rose, supra note 270 (manuscript at 35–44, 49–57) (evaluating SEC’s whistleblower bounty program and discussing possibility of qui tam safety valve). See generally David Freeman Engstrom, Whither Whistleblowing? Bounty Regimes, Regulatory Context, and the Challenge of Optimal Design, 15 Theoretical Inquiries L. 605, 608–09 (2014) [hereinafter Engstrom, Whither Whistleblowing] (summarizing literature). 2014] PRIVATE ENFORCEMENT’S PATHWAYS 2001 ambiguities rather than merely surfacing information about obviously illegal conduct.274 At the same time, however, the possibility that litigation regimes will exhibit “life cycles” of development suggests that one might worry less about statutory drift in the securities context. After all, the current system is now decades old and so arguably fully mature, leaving fewer legal ambiguities and thus fewer opportunities for private legal innovation than were arguably present following the FCA’s 1986 revival. One might be comparatively more worried about public control when choosing a qui tam system, or private enforcement more generally, to implement a newly minted regulatory regime. Of course, some caveats are in order. For one, applying empirical insights drawn from qui tam to other regulatory areas must confront standard concerns about generalizability. After all, the FCA is, in a number of key respects, unusual or even sui generis in its structure and subject matter.275 This is a limitation of any microinstitutional analysis.276 But note that the FCA regime is also in many ways a conservative test because of the DOJ’s unique, and uniquely strong, statutory gatekeeper powers, which exist nowhere else within the American regulatory state.277 To that extent, one might be more concerned about, and expect to find stronger evidence of, statutory “drift” and weak public control of litigation in the overwhelming majority of regulatory areas where agencies do not possess any formal litigation gatekeeper powers. Future research might thus extend the account of private enforcement’s pathways presented herein to regulatory areas beyond the FCA or Dodd-Frank, specifying how institutional structure and ground-level realities shape the pathways down which private enforcement travels. In addition, it should by now be apparent that firm conclusions regarding optimal regulatory design in a given regulatory area will not

274. See Engstrom, Whither Whistleblowing, supra note 273, at 619 (noting this possibility). 275. The main way the FCA is unusual, addressed more fully below, derives from the DOJ’s statutory gatekeeper powers. In addition, some might argue the qui tam regime is unique because it protects the government’s own fiscal interests as a purchaser, rather than protecting citizen interests as consumers or employees, thus giving the government a greater fiscal stake in the litigation. But it is not clear if this is a valid distinction, or why it should matter even if it were. After all, qui tam suits under the FCA also protect taxpayer interests. Nor is it the case that FCA enforcement agencies, such as the DOJ, have a direct fiscal stake in litigation outcomes, apart from the small portion of recoveries from successful healthcare enforcement actions the DOJ retains via the Health Care Fraud and Abuse Control Account established under the Health Insurance Portability and Accountability Act of 1996 (HIPAA). See Lemos & Minzner, supra note 18, at 864–65 (describing HIPAA revolving fund used by agencies to fund enforcement of healthcare- related federal law). 276. See Edward L. Rubin, Commentary, The New Legal Process, the Synthesis of Discourse, and the Microanalysis of Institutions, 109 Harv. L. Rev. 1393, 1437 (1996) (calling for “microanalysis of institutions” to drive theoretical development). 277. See Engstrom, Gatekeepers, supra note 16, at 646 tbl.2 (cataloguing existing litigation-oversight regimes). 2002 COLUMBIA LAW REVIEW [Vol. 114:1913 only depend on context-specific evidence about private enforcers, but also theory and evidence about the behavior of public enforcers. This is a useful reminder that, while the above analysis does much to specify and empirically test some of the core attributes of private enforcement, the use of private litigation as a regulatory tool is very much a choice over competing alternatives that have their own attributes.278 In making design decisions, regulatory architects will thus need to rely on growing scholarly work that seeks to understand the motives and means of public enforcers.279 Only through further theoretical and empirical development on both sides of the public–private divide can legal scholars generate a synthetic understanding of the legal and policy consequences of regulatory choices.

C. Litigation and Democracy A final set of insights, and a final scholarly contribution, concern the complex relationship between litigation and democracy. A long literature considers that relationship, focusing on, among other things, the allocation of power between judge and jury;280 the effect of the “vanishing trial,” private settlement, and alternative dispute mechanisms on the public elaboration of legal norms;281 the effect of procedures, particularly the class-action mechanism, on underlying substantive legal rights;282

278. See Neil K. Komesar, Imperfect Alternatives: Choosing Institutions in Law, Economics, and Public Policy 3–23 (1994) (noting all institutional analysis is inherently comparative). 279. For a recent theoretical synthesis on the public side of the public–private divide, see Lemos & Minzner, supra note 18, at 854 (arguing public enforcers’ “incentives are strongest when enforcement agencies are permitted to retain all or some of the proceeds of enforcement”); see also Nuno Garoupa & Daniel Klerman, Optimal Law Enforcement with a Rent-Seeking Government, 4 Am. L. & Econ. Rev. 116, 117 (2002) (modeling public enforcement as effort to maximally appropriate rents of illegal conduct). 280. See, e.g., Jeffrey Abramson, We, the Jury: The Jury System and the Ideal of Democracy 8 (1994) (arguing for “vision that defends the jury as a deliberative rather than a representative body”); David Millon, Juries, Judges, and Democracy, 18 Law & Soc. Inquiry 135, 147–57 (1993) (reviewing Shannon C. Stimson, The American Revolution in the Law: Anglo-American Jurisprudence Before John Marshall (1990)) (surveying history of debate over allocation of power between judge and jury); Judith Resnik, Managerial Judges, 96 Harv. L. Rev. 374, 386–414 (1984) (describing increasingly active judicial role in civil litigation, particularly at pretrial stage). 281. See, e.g., Stephen B. Burbank & Stephen N. Subrin, Litigation and Democracy: Restoring a Realistic Prospect of Trial, 46 Harv. C.R.-C.L. L. Rev. 399, 401 (2011) (asserting infrequency of trials can “diminish democracy”); Owen Fiss, Against Settlement, 93 Yale L.J. 1073, 1085 (1984) (arguing settlement diminishes “social function” of adjudication in which judges “explicate and give force to the values embodied in authoritative texts such as the Constitution and statutes”); Marc Galanter, The Vanishing Trial: An Examination of Trials and Related Matters in Federal and State Courts, 1 J. Empirical Legal Stud. 459, 459–60 (2004) (reviewing empirical evidence regarding declining frequency of trials). 282. See, e.g., supra note 13 (collecting books and articles by Martin Redish asserting aggregated litigation undermines democracy); see also Stephen B. Burbank, Procedure, 2014] PRIVATE ENFORCEMENT’S PATHWAYS 2003 and, more generally, the conditions under which litigation and judicial intervention improves or degrades democratic politics.283 Of course, one’s view on such issues will depend, often quite heavily, on one’s theory of democracy. Thus, legal scholars who hold a more formal, institution-focused theory of democratic politics tend to cast any lawmaking that occurs beyond legislative or administrative precincts as lacking a democratic pedigree.284 By contrast, scholars who adhere to deliberative or public-choice-inspired theories of democratic politics tend to conclude that litigation gives voice, and even influence, to those who have been excluded from legislative or executive decisionmaking.285 For those holding a public-choice view of politics, even unchecked, privately driven legal innovations of the sort Part I hypothesizes can improve, rather than degrade, democratic politics by offering a salutary

Politics, and Power: The Role of Congress, 79 Notre Dame L. Rev. 1677, 1710 (2004) (noting, following promulgation of Rule 23 and related developments, “it . . . became increasingly clear that federal courts wielded enormous power under the banner of procedure and that many choices they made under (or under authority of) Federal Rules had consequential substantive impact”); Paul D. Carrington, “Substance” and “Procedure” in the Rules Enabling Act, 1989 Duke L.J. 281, 282 (defending judicial rulemaking under Rules Enabling Act as consistent with democratic commitments because “complex technical issues of judicial practice cannot sustain attention through the political process”). 283. See, e.g., Gordon Silverstein, Law’s Allure: How Law Shapes, Constrains, Saves, and Kills Politics 3–5 (2009) (exploring costs and benefits of “juridification” of American politics and public policy); see also Stuart A. Scheingold, The Path of the Law in Political Science: De-centering Legality from Olden Times to the Day Before Yesterday, in The Oxford Handbook of Law and Politics 746–48 (Keith Whittington et al. eds., 2008) (cataloguing scholarly literatures rethinking litigation–democracy relationship by exploring “politicization of legality,” including use of litigation to secure electoral accountability, role of litigation as collective political mobilization, and courts as agents of institutional reform). For a popularized version of arguments about the litigation– democracy relationship, see Walter K. Olson, The Rule of Lawyers: How America’s New Litigation Elite Threatens Democracy 23 (2003) (examining rise of mass-tort class-action suits in late twentieth century). 284. This, in any event, seems to be an implicit assumption of Redish’s work. See, e.g., Redish, Wholesale Justice, supra note 13, at 3–15, 27 (espousing democratic theory grounded in “representation and accountability” and critiquing class actions as surreptitious judicial lawmaking outside of “traditionally recognized democratic procedures”). 285. See, e.g., Barnes, supra note 60, at 41 (“[A] flexible and politically responsive federal bench can promote democratic values, such as giving voice to interests that fail to participate in legislative or executive decision-making.”); M. Elizabeth Magill, Courts and Regulatory Capture, in Preventing Regulatory Capture, supra note 31, at 397, 408 (asserting courts are less vulnerable to capture-like dynamics). But see Galanter, Why the “Haves” Come Out Ahead, supra note 79, at 98–101, 103–04 (noting repeat-player advantage in litigation regimes); Edward L. Glaeser & Andrei Shleifer, The Rise of the Regulatory State, 41 J. Econ. Literature 401, 404–08 (2003) (arguing courts are more vulnerable than agencies to “subversion” by powerful interests). For overviews of public- choice theory and deliberative theories of democracy, see generally Debating Deliberative Democracy (James S. Fishkin & Peter Laslett eds., 2003); Jerry L. Mashaw, Greed, Chaos, and Governance: Using Public Choice to Improve Public Law (1997). 2004 COLUMBIA LAW REVIEW [Vol. 114:1913 counterweight to “capture” and other patterns of political control within the legislative or administrative process.286 A richer understanding of private enforcement’s pathways provides a foothold for thinking about these questions, and further debate along these lines would no doubt prove valuable. But exploring private enforcement’s pathways also makes a narrower, yet critically important, contribution to a relatively new entrant in the litigation-and-democracy debate: the “regulation by litigation” concept that animates much of the above discussion.287 Recent scholarship has gradually built up a critique of “regulatory litigation” against much of an industry as suffering from various pathologies, among them a lack of transparency and compre- hensiveness relative to the legislative- or administrative-rulemaking process.288 The above analysis has engaged, and in important ways qualified, such criticisms—hence the insight, noted previously, that an omnibus, pan-industry lawsuit might, contrary to qui tam critics’ concerns, be less worrying from a public-control perspective than multiple, serial litigation efforts.289 Yet a richer accounting of private enforcement’s pathways helps to address a deeper problem with the scholarly literature that has to this point evaded mention: The critique of “regulatory litigation” has for the most part been developed in the context of large-scale, and often sui generis, litigations, such as those against the gun, tobacco, and lead-paint industries.290 As a result, the theory lacks generalizable implications beyond relatively few litigation and policymaking episodes.

286. Alternatively, private legal innovation of the sort examined in this Article, which tends to expand statutory reach, may merely counteract the effect of repeat-play dynamics noted previously, see supra note 79 and accompanying text, which often favor defendants and so tend to limit statutory reach. As a concrete example, top practitioners within the relator and defense-side qui tam bars have expressed the belief that “Loser’s Rules,” as one of them put it, in which one-shotter and often less sophisticated relators and their counsel appeal bad cases that yield overly broad appellate rulings cabining FCA liability, are as important to the interpretative pathways down which the FCA has traveled as are pervasive relator-side efforts to expand the statute’s reach. See, e.g., Email from FCA Defense Counsel and Former Attorney, Civil Fraud Section, U.S. Dep’t of Justice (Sept. 3, 2014, 10:00 AM) (on file with Columbia Law Review). 287. See supra notes 70, 72, 186, 197 and accompanying text (discussing “regulation by litigation” at greater length). 288. See, e.g., Morriss et al., supra note 70, at 51 (arguing “regulation by litigation” is “potentially less effective form of regulation” because “regulation imposed through litigation is always less comprehensive than regulation imposed by rulemaking”). 289. See supra note 185–187 and accompanying text (discussing public-control concerns with omnibus lawsuits). 290. See, e.g., Martha A. Derthick, Up in Smoke: From Legislation to Litigation in Tobacco Politics 224 (3d ed. 2012) (“The tobacco settlement was public policy of a kind that the Framers of the Constitution tried to protect Americans against.”); Morriss et al., supra note 70, at 160–77 (analyzing litigation against diesel-engine, tobacco, and chemical companies and concluding, in those cases, “regulation-by-litigation . .. did not serve the public interest” and was not appropriate); Regulation Through Litigation, supra note 70, at 1–20 (surveying litigation against tobacco, lead-paint, breast-implant, gun, and HMO 2014] PRIVATE ENFORCEMENT’S PATHWAYS 2005

A more rigorous specification of the mechanisms that might drive private enforcement’s pathways begins to address this problem by suggesting that concerns about the increasingly “regulatory” cast of litigation efforts are not limited to the blockbuster litigations that have tended to attract popular and academic commentary. To the contrary, concerns about litigation’s increasing scale and regulatory scope and its control by politically accountable actors may be deeply embedded in the structure of private enforcement itself. In this way, the account of private enforcement’s pathways offered here extends thinking about “regulatory litigation” beyond marquee litigation like the tobacco cases to the more prosaic, workaday private enforcement regimes, from job discrimination and securities to antitrust and consumer protection, that dot the American regulatory landscape.

CONCLUSION

Legal scholars have spilled barrels of ink assessing whether politically accountable actors can exercise meaningful public control over bureaucratic implementers of public policy. But few studies consider public control of implementers of no less import in the modern American regulatory state: private litigants deputized as private attorneys general across a wide range of regulatory contexts. This Article has attempted to make a start, however haltingly, down that road by sketching and then empirically testing competing accounts of private enforcement’s pathways. The most basic aim has been to push past relatively simplistic models of private enforcement and advance a subtler understanding of the legal industries and finding “policies that result from litigation almost invariably involve less public input and accountability than government regulation”); Silverstein, supra note 283, at 246–64 (“[T]he case of the antitobacco advocates does provide an illustration of how law can save politics—and how law can kill politics as well.”). At least one commentator refers to these cases as “mega social policy cases,” explicitly distinguishing them from more traditional “social impact litigation and . . . class actions.” See Michael L. Rustad, Smoke Signals from Private Attorneys General in Mega Social Policy Cases, 51 DePaul L. Rev. 511, 511, 514 (2001). Interestingly, an earlier profusion of scholarly debate about the relationship between litigation and democracy is subject to a similar criticism of its focus on large-scale institutional-reform litigation. For classic accounts, see Malcolm M. Feeley & Edward L. Rubin, Judicial Policy Making and the Modern State: How the Courts Reformed America’s Prisons 351–60 (2000) (focusing on “national character” of prison reform and analyzing role of judicial policymaking in political process); Ross Sandler & David Schoenbrod, Democracy by Decree: What Happens When Courts Run Government 98– 112 (2003) (discussing development of “soft rights” and enforcement of such rights in “institutional reform cases”); Abram Chayes, The Role of the Judge in Public Law Litigation, 89 Harv. L. Rev. 1281, 1281–84 (1976) (examining place of “public law,” such as “school desegregation, employment discrimination, and prisoners’ or inmates’ rights cases” in “American political and legal system”). For a rare contrary view, see Thomas F. Burke, The Rights Revolution Continues: Why New Rights Are Born (and Old Rights Rarely Die), 33 Conn. L. Rev. 1259, 1259 (2001) (“Regulation by litigation is, then, best understood not as a novelty, but rather as an extension of tendencies characteristic of American public policy.”). 2006 COLUMBIA LAW REVIEW [Vol. 114:1913 and policy consequences of the legislative deployment of private litigation as a regulatory tool. Much work remains to be done to widen the inquiry begun here to other contexts. Only by doing so can one achieve a fuller specification of the institutional and other conditions under which private litigation poses greater or lesser calibration and control challenges. This Article hopes to motivate and guide that work while moving toward a more synthetic understanding of the choice of private enforcement over regulatory alternatives.